CN102737059B - For determining the method for the accuracy information of resource description information, device and equipment - Google Patents
For determining the method for the accuracy information of resource description information, device and equipment Download PDFInfo
- Publication number
- CN102737059B CN102737059B CN201110093719.0A CN201110093719A CN102737059B CN 102737059 B CN102737059 B CN 102737059B CN 201110093719 A CN201110093719 A CN 201110093719A CN 102737059 B CN102737059 B CN 102737059B
- Authority
- CN
- China
- Prior art keywords
- resource
- description information
- resource description
- information
- key word
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 230000000875 corresponding Effects 0.000 claims description 65
- 238000000034 method Methods 0.000 claims description 21
- 238000010276 construction Methods 0.000 claims description 5
- 239000004744 fabric Substances 0.000 claims 1
- 238000010586 diagram Methods 0.000 description 8
- 230000000694 effects Effects 0.000 description 6
- 230000003542 behavioural Effects 0.000 description 4
- 230000015572 biosynthetic process Effects 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 4
- 238000005755 formation reaction Methods 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 230000000052 comparative effect Effects 0.000 description 2
- 230000002596 correlated Effects 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 238000000926 separation method Methods 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 239000012141 concentrate Substances 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
Abstract
The present invention provides method, device and the equipment of a kind of accuracy information for determining resource description information.Pending resource description information is selected according in multiple resource description information that the solution of the present invention is first comprised by pre-established resource description information set;, then obtain the distributed intelligence in other resource description information described of each key word that described pending resource description information comprised then;Subsequently according to described distributed intelligence, determine the degree of association between described pending resource description information and/or its each key word comprised and every other resource description information, to obtain the accuracy information of this pending resource description information.The invention have the advantages that the accuracy that can determine resource description information to the description of resource.
Description
Technical field
The present invention relates to computer realm, particularly relate to a kind of accuracy information determining resource description information method,
Device and equipment.
Background technology
Along with popularizing of network, the resource that increasing user hankers after marking oneself (is also referred to as UGC money
Source) issued by network, in order to share with other people.But, owing to individual subscriber has randomness to the mark of resource, often
Often arbitrarily can mark according to personalized preference, emotion etc., the accuracy of the information therefore marked is difficult to ensure that.Such as, user
After trick the picture of A star being labeled as B star, issue the photograph album at oneself subsequently and concentrate.Then pass through as other users
During search engine search B star, the picture of A star possibly be present in Search Results, thus have a strong impact on search engine can
Reliability.
Summary of the invention
It is an object of the invention to provide method, device and the equipment of a kind of accuracy information determining resource description information.
According to an aspect of the present invention, it is provided that a kind of computer implemented accuracy for determining resource description information
The method of information, wherein, the method comprises the following steps:
Multiple resource description information that a is comprised by pre-established resource description information set select pending resource to retouch
Stating information, wherein, each resource description information in the plurality of resource description information is used to describe a resource, and each
Described by resource described by resource description information and other resource description information arbitrary in this resource description information set
Resource is similar or identical;
Each key word that the described pending resource description information of b acquisition is comprised is in other resource description information described
Distributed intelligence;
C according to described distributed intelligence, determine described pending resource description information and/or its each key word comprised with
The degree of association between every other resource description information, to obtain the accuracy information of this pending resource description information.
According to another aspect of the present invention, additionally provide a kind of computer implemented for determining the accurate of description information
The accuracy of degree information determines device, and wherein, this accuracy determines that device includes:
Select device, select in the multiple resource description information comprised by pre-established resource description information set
Pending resource description information, wherein, each resource description information in the plurality of resource description information is used to describe one
Other resource descriptions arbitrary in individual resource, and the resource described by each resource description information and this resource description information set
Resource described by information is similar or identical;
First acquisition device, for obtain each key word that described pending resource description information comprised described its
Distributed intelligence in his resource description information;
First determine device, according to described distributed intelligence, determine described pending resource description information and/or its comprise
The degree of association between each key word and every other resource description information, to obtain the accuracy of this pending resource description information
Information.
According to a further aspect of the invention, a kind of computer equipment is additionally provided, wherein, before this computer equipment includes
State accuracy and determine device.
Compared with prior art, the invention have the advantages that 1) can be by the resource description information to a resource
The key word comprised distribution situation in the resource description information of other multiple same or similar resources, determines that this resource is retouched
State the degree of association of information or its key word comprised and other resource description information, due to the money described by this resource description information
Source and the resource described by other resource description information are same or similar, and therefore, this degree of association can reflect that this resource description is believed
Breath or the description accuracy of its key word comprised, particularly user generate the description accuracy of the resource description information of resource;
2) by key word that pending resource description information is comprised in the description information of other multiple same or similar resources
Distribution situation and the analysis of other relevant informations, it is possible to more accurately determine pending resource description information and/or it comprises
Key word and other resource description information between the degree of association, thus more precisely judge the standard of pending resource description information
Exactness;3) can by determined by the accuracy of resource description information be applied to multiple occasion, such as: a) be applied to retrieval system
System, so that the inaccurate resource of resource description information sorts rearward, the sequence making retrieval result is the most reasonable;B) it is applied to recommend
System, such as, based on determined by the accuracy of resource description information recommend resource to user, to improve the utilization of resource
Rate;C) prompt system, such as, based on determined by the accuracy of resource description information point out the description of this resource of user may
Accuracy is relatively low.
Accompanying drawing explanation
By the detailed description that non-limiting example is made made with reference to the following drawings of reading, other of the present invention
Feature, purpose and advantage will become more apparent upon:
Fig. 1 is the flow chart of the method for the accuracy information for determining resource description information of one aspect of the invention;
Fig. 2 is the flow process carrying out pre-established resource description information set based on resource cluster of a preferred embodiment of the invention
Figure;
Fig. 3 is the stream of the method for the accuracy information for determining resource description information of a preferred embodiment of the invention
Cheng Tu;
Fig. 4 be a preferred embodiment of the invention according to determined by the accuracy information of resource description information come money
Source performs the flow chart of corresponding operating;
Fig. 5 is that the accuracy of the accuracy information for determining resource description information of one aspect of the invention determines device
Schematic diagram;
Fig. 6 is the based on the next pre-established resource description information set of resource cluster accurate of a preferred embodiment of the invention
Degree determines device schematic diagram
Fig. 7 is that the accuracy of the accuracy information for determining resource description information of a preferred embodiment of the invention is true
Determine device schematic diagram;
Fig. 8 be a preferred embodiment of the invention according to determined by the accuracy information of resource description information come money
Source performs the accuracy of corresponding operating and determines device schematic diagram;
In accompanying drawing, same or analogous reference represents same or analogous parts.
Detailed description of the invention
Below in conjunction with the accompanying drawings the present invention is described in further detail.
Fig. 1 shows the flow process of the method for the accuracy information for determining resource description information of one aspect of the invention
Figure.Wherein, the method according to the invention is mainly completed by the operating system in computer equipment or processing controller, for letter
See from tomorrow, below described operating system or processing controller are referred to as accuracy and determine device.Wherein, this computer equipment bag
Include but be not limited to: 1) subscriber equipment;2) network equipment.Described subscriber equipment includes but not limited to computer, smart mobile phone, PDA
Deng;The described network equipment include but not limited to server group that single network server, multiple webserver form or based on
The cloud being made up of a large amount of computers or the webserver of cloud computing (Cloud Computing), wherein, cloud computing is distributed
The one calculated, the super virtual machine being made up of a group loosely-coupled computer collection.
In step sl, described accuracy determines that device multiple is retouched by what pre-established resource description information set comprised
State and information selects pending resource description information, wherein, each resource description information in the plurality of resource description information
It is used to describe a resource, and arbitrary with this resource description information set of the resource described by each resource description information
Resource described by other resource description information is similar or identical.Wherein, described resource includes but not limited to: 1) picture category money
Source;2) audio class resource;3) video class resource;4) program bag class resource etc..
Wherein, the mode of pre-established resource description information set includes but not limited to:
1) artificial next pre-established resource description information set.
A) for picture category resource, operator when setting up resource description information set, view-based access control model effect judges
Multiple resources are the most same or similar.Such as, if for picture category resource resource A1 and resource B 1 phase in visual effect
With, being only merely and there are differences at aspects such as background color, size, regional areas, then operator judge resource A1 and resource B
1 is similar.
B) for video class resource, operator, when setting up resource description information set, judge based on resource plot
Multiple resources are the most same or similar.Such as, if resource A2 is identical with the main action of resource B2, simply in image resolution
The aspect such as rate, compressed format is different, then operator judge that resource A2 is similar to resource B2.
C) for audio class resource, operator, when setting up resource description information set, judge based on auditory effect
Multiple resources are the most same or similar.Such as, resource A3 and resource B3 are identical on auditory effect, be different only in that resource A3 with
The aspects such as the lyrics of resource B3, compressed format are different, then operator judge that resource A3 is similar to resource B3.
D) for program bag class resource, based on program source code, operator judge that multiple resource is the most identical or phase
Seemingly.Such as, resource A4 and the source code of resource B4 are simply in the name of variable, pointer, array etc. or to program source code
There is difference in the aspects such as explanation, then operator judge that resource A4 is similar to resource B4.
2) pre-established resource description information set is carried out based on resource cluster.This sets up mode will in the embodiment depicted in figure 2
Described in detail.
It should be noted that the example above is only better described technical scheme, rather than the limit to the present invention
System, it should be appreciated by those skilled in the art that and any determine the most same or analogous mode of resource, and based on same or similar
Resource carrys out the implementation of pre-established resource description information set, should be included in the scope of the present invention.
Described accuracy determines in multiple description information that device is comprised by pre-established resource description information set to be selected
The selection mode selecting pending resource description information includes but not limited to:
1) pending resource description information is randomly choosed.
2) described accuracy determines that device is according to network phase corresponding to the resource described by the plurality of resource description information
Pass information, is identified the plurality of resource description information, using by identify gained user generate resource description information as
Described pending resource description information.
Such as, the plurality of resource description information includes the resource description information of resource A from website A ', from website
The resource description information of resource B of B ' and the resource description information of resource C from website C ', described accuracy determines device root
Determining that described website A ' and website B ' is authoritative website according to predetermined authoritative website list, website C ' is inauthoritativeness website, therefore,
Described accuracy determines that device is retouched according to the resource of the resource description information of resource A, the resource description information of resource B and resource C
State information from website, identify the resource description information of resource C from inauthoritativeness website, and by the resource description of resource C
Information is as described pending resource description information.
Preferably, described network related information include following at least one:
A) link address information of the resource that this network related information is corresponding.Specifically, described accuracy determines device root
According to the pre-determined text information comprised in the link address information of resource, such as: i) bbs;ii)blog;Iii) SNS etc., identify
Resource description information corresponding to this resource is that user generates resource description information, and then by resource description information corresponding for this resource
As pending resource description information.Such as, the resource described by the plurality of resource description information includes resource A and resource B.
Wherein, the link address information of resource A is " www.222.com ", and the link address information of resource B is " bbs.444.com ", then
Described accuracy determines that device comprises " bbs " according to the link address information of resource B, identifies that the resource description information of resource B is
User generates resource description information, and using the resource description information of resource B as pending resource description information.
B) the page feature information of webpage belonging to the resource that this network related information is corresponding.Specifically, described accuracy is true
Determine device according to the code of webpage belonging to resource being analyzed the page feature information of gained, such as, model category feature information,
The particular text information etc. such as such as " blog " that be contained in page subject matter, " album ", determine and belong to this webpage
The resource description information of resource be that user generates resource description information, and then using resource description information corresponding for this resource as
Pending resource description information.Preferably, described model category feature information includes;1) model such as " main building ", " 1st floor ", " building-owner "
Class text information;2) the model class formation information of the display module etc. that multiple stacking shows and structure is identical is comprised.
C) the page feature information of the webpage that the resource affiliated web site that this network related information is corresponding is comprised.Specifically,
Described accuracy determines that device is analyzed the net of this website of gained according to the web page code being comprised resource affiliated web site
Page page feature information, such as, occurs in the spies such as such as " blog " in the page subject matter of multiple webpage, " home videos "
The model class formation information etc. determine text message, occurring in multiple webpage, determines that the resource of the resource belonging to this webpage is retouched
The information of stating is that user generates resource description information, and then using resource description information corresponding for this resource as pending resource description
Information.
It is highly preferred that accuracy determines that device, according at least one in above-mentioned three network related information, comes many to this
Individual resource description information is identified, and generates resource description information using the user by identification gained and retouches as described pending resource
State information.
Such as, determine that device obtains the pre-determined text information " bbs " comprised in the link address information of resource when accuracy
Time, whether the page feature information of webpage belonging to resource of analyzing further comprises model category feature information, and when page feature is believed
When breath comprises model category feature information, this resource identification is that user generates resource description information by, and using this resource as institute
State pending resource description information.
It should be noted that the example above is only better described technical scheme, rather than the limit to the present invention
System, it should be appreciated by those skilled in the art that any can be for identifying that resource generates the net of resource description information to obtain user
Network relevant information, should be included in the scope of the present invention.
Then, in step s 2, described accuracy determines that device obtains what described pending resource description information was comprised
The distributed intelligence in other resource description information described of each key word.Wherein, when described pending resource description information only
When comprising the key word of one or more separation, described accuracy determines that device directly obtains the one or more key word and exists
Distributed intelligence in other resource description information described;When described pending resource description information comprises one or more text
Time, described accuracy determines that the one or more text is cut the process such as word, duplicate removal and obtained described pending by device
The key word that resource description information comprises.
Wherein, described distributed intelligence include following at least one:
1) total degree that each key word described occurs in described every other resource description information.Such as, pre-established
Resource description information set include describe resource A pending resource description information a, describe resource B resource description letter
Breath b and resource description information c of description resource C, described accuracy determines that device obtains the key word that resource description information a comprises
Including key word a1 and key word a2 and obtain key word a1 and key word a2 and occur in resource description information b 2 times, in money
Occurring 1 time in Source Description information c, the most described accuracy determines that device obtains the resource description information a bag describing pending resource
The total degree that the key word a1 contained and key word a2 occurs in described resource description information b with resource description information c is 2+1=
3 times.
2) number of times that each key word described occurs in described every other resource description information respectively.Such as, build in advance
Vertical resource description information set includes describing pending resource description information d of resource D, describing the resource description of resource E
Information e and resource description information f describing resource F;Described accuracy determines that device obtains pending resource description information d and comprises
Key word include key word d1 and key word d2 and obtain key word d1 in resource description information e and resource description information f
Occur 5 times, obtain key word d2 and occur 3 times in resource description information e with resource description information f.
3) identification information of other resource description information described of arbitrary key word in each key word described is comprised.Example
As, pre-established resource description information set includes describing pending resource description information g of resource G, describing the money of resource H
Source Description information h and resource description information i describing resource I;Described accuracy determines that device obtains pending resource description letter
The key word that breath g comprises includes key word g1 and determines that resource description information h comprises key word g1, in resource description information i not
Comprising key word g1, it is h that the most described accuracy determines that device obtains the identification information of the resource description information comprising key word g1.
4) quantity of other resource description information described of at least one key word described is comprised.Such as, pre-established money
Source Description information aggregate include describe resource J pending resource description information j, describe resource K resource description information k with
Describing the resource description information 1 of resource L, described accuracy determines that device obtains the key that pending resource description information j comprises
Word includes key word j1 and key word j2 and determines and comprise key word j1 in resource description information k, wraps in resource description information 1
Containing key word j2, the most described accuracy determine device obtain comprise in key word j1 and key word j2 at least one described other
The quantity of resource description information is 2.
5) quantity of other resource description information described comprising at least one key word described accounts for described all resources and retouches
State the ratio of the quantity of information.
6) quantity of other resource description information that each key word in each key word described is occurred accounts for all money
The ratio of the quantity of Source Description information.Such as, a key word occurs in 4 other resource description information, and all resources
The quantity of description information is 10, then this key word accounts for the quantity of all resource description information in the quantity of other resource description information
Ratio be 0.4.
Then, in step s 4, described accuracy determines that device, according to described distributed intelligence, determines described pending resource
The degree of association between description information and/or its each key word comprised and every other resource description information, waits to locate obtaining this
The accuracy information of reason resource description information.
Wherein, described accuracy determine device according to described distributed intelligence, determine described pending resource description information and/
Or the mode of the degree of association between its each key word comprised and every other resource description information includes but not limited to:
1) directly using distributed intelligence as described pending resource description information and/or its each key word comprised and institute
There is the degree of association between other resource description information.Such as, described accuracy determines that device acquisition comprises described pending resource
The quantity of other resource description information described of at least one key word accounts for the ratio of the quantity of described all resource description information
Being 0.8, the most described accuracy determines that device determines between described pending resource description information and every other resource description information
The degree of association be 0.8.The most such as, the number of other resource description information that each key word in each key word described is occurred
The ratio of the quantity that amount accounts for all resource description information is 0.4, and the most described accuracy determines that device determines that this key word is with all
The degree of association between other resource description information is 0.4.
2) carry out distributed intelligence processing obtained result as described pending resource description information and/or its
The degree of association between each key word comprised and every other resource description information.Specifically, carry out distributed intelligence processing
Mode includes: a) obtain the described degree of association according in distributed intelligence, such as: i) distributed intelligence entered with predetermined threshold
Row compares, and determines described pending resource description information and/or its each key word comprised and institute according to comparative result
There is the relevance level between other resource description information;Ii) ask for distributed intelligence to retouch with the resource in resource description information set
State the ratio of information sum, and determine described pending resource description information and/or its each pass comprised according to gained ratio
The degree of association between keyword and every other resource description information;B) described pending money is obtained according to multinomial in distributed intelligence
The degree of association between Source Description information and/or its each key word comprised and every other resource description information, such as: i) by two
Item distributed intelligence be used for described pending resource description information and/or its each key word comprised and every other money
The degree of association between Source Description information;Ii) multinomial distribution information is normalized, and the value of normalized gained is entered
Row sue for peace, average, ask logarithm and etc. process, using the value of gained as described pending resource description information and/or its
The degree of association between each key word comprised and every other resource description information;Iii) come multinomial distribution according to predetermined formula
Information carries out calculation process, and using the value of calculation process gained as described pending resource description information and/or its comprise
The degree of association etc. between each key word and every other resource description information.
Such as, described accuracy determines that each key word of the device pending resource of acquisition is retouched in described every other resource
Stating the total degree occurred in information is 10 times, and the most described accuracy determines that device is higher than the first predetermined threshold based on this total degree,
Determine that the degree of association between described pending resource description information and every other resource description information is senior.
The most such as, described accuracy determines that the key word that the pending resource description information of device acquisition comprises owns described
The number of times occurred in other resource description information is 5 times, and the most described accuracy determines that device makes a reservation for less than second based on this number of times
Threshold value, determines that the degree of association between described pending resource description information and every other resource description information is rudimentary.
The most such as, described accuracy determines that device acquisition comprises the key word Y that described pending resource description information comprises
The quantity of other resource description information described be 6, and obtain key word X, pass that described pending resource description information comprises
The total degree that keyword Y and key word Z occurs in described every other resource description information is 60 times, and the most described accuracy determines
Device will comprise the quantity of other resource description information described in this key word Y with each key word in described every other resource
The ratio 6/60=0.1 of the total degree occurred in description information is as between this key word Y and every other resource description information
The degree of association.
The most such as, described accuracy determines that device obtains each key word of pending resource in described every other resource
The total degree occurred in description information is 20 times, and based on comprising other moneys described in arbitrary key word in each key word described
The identification information of Source Description information obtains and comprises other resource description information described of arbitrary key word in each key word described
Quantity be 5, the most described accuracy determine device by each key word of described pending resource in described every other resource
In description information, the total degree of appearance is retouched with other resources described of arbitrary key word in each key word described that comprise of acquisition
State the ratio of number 20/5=4 of information as the pass between described pending resource description information and every other resource description information
Connection degree.
The most such as, described accuracy determines that device obtains each key word of pending resource in described every other resource
The total degree occurred in description information is 10 times, and in described resource description information set, all of key word quantity is 50, comprises
The quantity of other resource description information described of described arbitrary key word accounts for the ratio of the quantity of described all resource description information
Be 0.5, then accuracy determines that device is according to the first predetermined formula: described pending resource description information is retouched with every other resource
That states that each key word of the degree of association between information=pending resource occurs in described every other resource description information is total
Other resource descriptions described of all of key word quantity+described arbitrary key word in number of times/described resource description information set
The quantity of information accounts for the ratio of the quantity of described all resource description information, determines described pending resource description information and institute
There is the degree of association=10/50+0.5=0.7 between other resource description information.
The most such as, described accuracy determines that device obtains the key word V that comprises of pending resource and key word W in described institute
Having the total degree occurred in other resource description information is 10 times, it is thus achieved that the key word V that pending resource comprises is described all
The number of times occurred in other resource description information is 3 times, comprises at least one key word in described key word V and key word W
The ratio of the quantity that the quantity of other resource description information described accounts for described all resource description information is 0.9, then accuracy is true
Determine device according to the second predetermined formula: the key word that described pending resource description information comprises is believed with every other resource description
Number of times/pending resource that the degree of association between breath=this key word occurs in described every other resource description information comprises
The total degree * that each key word occurs in described every other resource description information comprises described at least one key word described
The quantity of other resource description information accounts for the ratio of the quantity of described all resource description information, determines that key word V is with all
The degree of association=3/10*0.9=0.27 between other resource description information.
It should be noted that the example above is only better described technical scheme, rather than the limit to the present invention
System, it should be appreciated by those skilled in the art that any according to described distributed intelligence, determine described pending resource description information and/
Or the implementation of the degree of association between its each key word comprised and every other resource description information, should be included in this
In bright scope.
Wherein, described accuracy information include following at least one: 1) described pending resource description information overall accurate
Exactness;2) accuracy of each key word that described pending resource description information is comprised.
Specifically, described accuracy determine device based on determined by pending resource description information and every other resource
The mode of the overall accuracy that the degree of association between description information obtains pending resource description information includes but not limited to: 1) straight
Connect the degree of association between described pending resource description information and every other resource description information as pending resource description
The overall accuracy of information.2) degree of association between described pending resource description information and every other resource description information is entered
The result that row process is obtained is as the overall accuracy of pending resource description information.Such as, by described pending money
The degree of association between Source Description information and every other resource description information is retouched as pending resource with the product of predefined weight value
State the overall accuracy of information.The most such as, by between described pending resource description information and every other resource description information
Each key word that the degree of association is asked for square or the result of 3 powers is comprised as described pending resource description information accurate
Degree etc..3) degree of association between each key word described pending resource comprised and every other resource description information is asked
The result obtained with summation after, weighted sum, quadrature, normalization etc. as pending resource description information overall accurately
Degree.
Described accuracy determines each key word and every other resource that device comprised based on described pending resource
The degree of association between description information determines the side of the accuracy of each key word that described pending resource description information comprised
Formula includes but not limited to;1) each key word directly described pending resource description information comprised and every other resource
Each key word that each degree of association between description information is comprised respectively as described pending resource description information accurate
Degree.2) each key word that described pending resource description information is comprised and associating between every other resource description information
Degree carries out processing each key that each result obtained is comprised respectively as described pending resource description information
The accuracy of word.Such as, between each key word described pending resource comprised and every other resource description information
Each degree of association is asked for square respectively or each results of 3 powers is comprised as described pending resource description information each
The accuracy of key word.The most such as, each key word described pending resource comprised is believed with every other resource description
Each pass that the product of each degree of association between breath and predefined weight is comprised respectively as described pending resource description information
The accuracy of keyword.
It should be noted that the example above is only better described technical scheme, rather than the limit to the present invention
System, it should be appreciated by those skilled in the art that any based on determined by the degree of association obtain the total of pending resource description information
The implementation of the accuracy of each key word that body accuracy and/or described pending resource description information are comprised, all should
Within the scope of the present invention.
The method according to the invention can be by the key word that comprises the resource description information of a resource more than other
Distribution situation in the resource description information of individual same or similar resource, determines this resource description information or its key comprised
Word and the degree of association of other resource description information, the resource described by this resource description information and other resource description information
Described resource is same or similar, and therefore, this degree of association can reflect this resource description information or its key word comprised
Accuracy is described.The method according to the invention is particularly suited for determining that the description of the resource description information that user generates resource is accurate
Degree.
One of preferred version as the present invention, Fig. 2 show a preferred embodiment of the invention based on resource cluster
Carry out the flow chart of pre-established resource description information set.
In step s 5, described accuracy determines that device obtains multiple resources.Wherein, described accuracy determines that device obtains
The mode of multiple resources includes but not limited to: 1) by obtaining the plurality of resource in multiple websites;2) by the resources bank of pre-stored
The middle the plurality of resource of acquisition etc..
It should be noted that the example above is only better described technical scheme, rather than the limit to the present invention
System, it should be appreciated by those skilled in the art that the implementation of the multiple resource of any acquisition, should be included in the scope of the present invention
In.
Then, in step s 6, described accuracy determines the information that device self is comprised according to the plurality of resource, comes
Clustering the plurality of resource, to obtain one or more groups cluster resource, wherein, often group cluster resource includes one or more
Same or analogous resource.Wherein, described accuracy determines that device uses corresponding cluster mode according to resource type.Example
As, for picture category resource, described accuracy determines picture element information, the color histogram of picture that device comprises according to picture
Information, local invariant feature (SIFT, Scale-invariant feature transform), textural characteristics (HTD,
Homogeneous Texture Descriptor), color characteristic (SCD) etc., carry out picture cluster.The most such as, for regarding
Frequently class resource, described accuracy determines that device enters according to the size of video resource, form, the information such as sectional drawing of identical time point
Row cluster.The most such as, for audio class resource, described accuracy determines that device is according to the form of audio frequency, size, audio resource
The information such as average pitch, the audio resource tone on each time point cluster.The most such as, program bag class is provided
Source, described accuracy determines that the source code information etc. that device comprises according to program bag clusters.
It should be noted that the example above is only better described technical scheme, rather than the limit to the present invention
System, it should be appreciated by those skilled in the art that and any resource clusters to obtain one or more groups cluster resource, wherein, often
Group cluster resource includes the cluster mode of one or more same or analogous resource, should be included in the scope of the present invention.
Then, in the step s 7, described accuracy determines device often to organize same or analogous resource corresponding according to described
Resource description information, sets up described resource description information set.
Such as, described accuracy determines that device obtains one group of cluster resource A1, one group of cluster resource A2, one group of cluster resource
A3, described accuracy determines that device is corresponding according to resource description information corresponding to cluster resource a1 that comprises of resource A1, resource a2
Resource description information and resource description information corresponding to resource a3, set up described resource description information set.
Preferably, before step S7, afterwards or simultaneously, described accuracy determines that device is based on cluster resource A2 or poly-
The resource description information that resource that class resource A3 comprises is corresponding, sets up another resource description information set.
Fig. 3 shows the method for the accuracy information for determining resource description information of a preferred embodiment of the invention
Flow chart.Wherein, step S1 is described in detail the most with reference to the embodiment shown in FIG. 1 with S2, and is contained in by reference
This, repeat no more.
In step s3, described accuracy determines that device obtains for other relevant informations determining the described degree of association.
Wherein, other relevant informations described include following at least one;
1) resource described by other resource description information described of arbitrary key word in each key word described is comprised
Authoritative.
Wherein, described accuracy determines that the authoritative mode of device acquisition resource includes but not limited to: a) obtains and prestores
The authority of this resource of storage;B) characteristic information based on this resource affiliated web site determines the authority of this resource.Such as, institute
State accuracy and determine whether device visit capacity based on this website, this website are included in predetermined authoritative website, material website
The quantity of the resource from this website comprised in list, in information bank whether exceed predetermined threshold and information bank comprise come
Whether it is high-quality etc. from the quality information of the resource of this website, determines the authority of this resource.
2) each key word in described all key words and comprise this key word each other resource description information between
The first degree of association.
Wherein, described accuracy determines that device obtains key word and each other resource description information comprising this key word
Between the mode of the first degree of association include but not limited to:
A) key word obtaining pre-stored is relevant to first between each other resource description information comprising this key word
Degree;Such as, other resource description information including key word X are resource description information b and resource description information c, and described
Accuracy determines that in the storage device that device can be accessed by, pre-stored key word X is relevant to first between resource description information b
Degree is 2, and the first degree of association between key word X and resource description information c is 3, then accuracy determines that device obtains the pass of pre-stored
Keyword X and comprise this key word other resource description information b and c between the first degree of association be respectively 2 and 3.
B) described accuracy determine device based on following at least one determine key word and comprise this key word one
Described first degree of association between other resource description information, with determine this key word respectively with comprise this key word each other
The first degree of association between resource description information:
I) number of times that this key word occurs in other resource description information;Such as, described accuracy determines device
The key that the number of times occurred in other resource description information by this key word and these other resource description information are comprised
The ratio of word sum, as the first degree of association between this key word and this other resource description information.
Ii) text type of the text message at this key word place;Wherein, described text message is contained in other resources and retouches
State in information, and described text type include but not limited to: title class text, Anchor Text class text, at webpage belonging to this resource
In the context class text etc. adjacent with resource;Such as, when the text type comprising this key word is title class text, then described
Accuracy determines that device determines that the first degree of association of this key word is senior.
Iii) number of times that this key word occurs in each text type that other resource description information comprise respectively
And the predefined weight value of each text type;Such as, described accuracy determines that device obtains this key word and retouches in these other resources
The title class text that the information of stating comprises occurs 1 time occur 8 times in context class text, and the predefined weight of title class text
Value is 0.6, and the predefined weight value of context class text is 0.3, and the most described accuracy determines that device determines different text type
Predefined weight value and this key word occur in the sum of products=0.6*1+0.3*8=3 of the number of times in dissimilar text, and should
Sum of products as this key word and comprise this key word these other resource description information between the first degree of association.
It should be noted that the example above is only better described technical scheme, rather than the limit to the present invention
System, it should be appreciated by those skilled in the art that any these other resource descriptions for determining this key word with comprise this key word
The implementation of the first degree of association between information, such as, the number of times that key word is occurred in other resource description information
It is multiplied by the meansigma methods of the predefined weight value of each text type of the text message at this key word place, obtains described first phase
Guan Du etc., should be included in the scope of the present invention.
3) each key word in described all key words and the second degree of association between described pending resource description information.
Wherein, described accuracy determines that each key word that device obtains in described all key words is believed with described pending resource description
With described accuracy, the acquisition mode of the second degree of association between breath, determines that device obtains each key in described all key words
Word and comprise this key word other resource description information between the acquisition mode of the first degree of association same or similar, and to quote
Mode be incorporated herein, repeat no more.
It should be noted that the example above is only better described technical scheme, rather than the limit to the present invention
System, it should be appreciated by those skilled in the art that any other relevant informations for determining the described degree of association and any acquisition are used
In the implementation of other relevant informations determining the described degree of association, should be included in the scope of the present invention.
Need it is further noted that step S2 and step S3 there is no sequencing.
Then, in step S4 ' in, described accuracy determines that device is according to described distributed intelligence and other relevant letters described
Breath, determines the degree of association between described pending resource description information and every other resource description information, pending to obtain this
The accuracy of resource description information.Wherein, described accuracy determines that device is correlated with according to described distributed intelligence and described other
Information, determine the mode of the degree of association between described pending resource description information and every other resource description information include but not
It is limited to:
1) described accuracy determines that device first determines other moneys comprising at least one key word based on described distributed intelligence
Source Description information, further according to determined by comprise the every other resource description information of at least one key word and other phases described
Pass information determines that described pending resource description information and/or its each key word comprised are believed with every other resource description
The degree of association between breath.Such as, described accuracy determines that device is first based on comprising at least one key word in each key word described
The identification information of other resource description information described, determine resource description set comprises at least one key word described its
His resource description information includes resource description information a describing resource A, and then, described accuracy determines that device is further according to resource A
Authority be senior, determine that the degree of association between described pending resource description information and every other resource description information is for height
Level.
The most such as, described accuracy determines that device is first based on comprising the institute of at least one key word in each key word described
State the identification information of other resource description information, determine other moneys described comprising at least one key word in resource description set
Source Description information includes resource description information b describing resource B and describes resource description information c of resource C, and determines description money
Resource description information b of source B comprises key word Y, and resource description information c describing resource C comprises key word X and key word Y, institute
State accuracy and determine that device the first degree of association based on key word X Yu resource description information c is 0.6, determine this key word X with
The degree of association between every other resource description information is 0.6, and the first degree of association based on key word Y Yu resource description information b
It is 0.8 and the first degree of association of key word Y and resource description information c is 0.4, determines this key word Y and every other resource
The degree of association=0.8+0.4=1.2 between description information.
2) accuracy determine device according at least one in distributed intelligence and other relevant informations described at least
One determines the described degree of association.Specifically, to determine that device adjusts based on other relevant informations described described for described accuracy
The value that distributed intelligence is comprised, and based on adjust after result determine described pending resource description information and/or its comprise
Each key word and every other resource description information between the degree of association.
Such as, described accuracy determines that device obtains key word X resource description information a in resource description information set
Middle appearance 2 times, and the first degree of association obtaining this key word X and resource description information a is 0.6, the most described accuracy determines dress
Put with this first degree of association as Dynamic gene, determine the described degree of association between this key word X and every other resource description information
For 0.6*2=1.2.
The most such as, described accuracy determines that device obtains key word Y resource description letter in resource description information set
Occurring 3 times in breath b, the first degree of association of key word Y and resource description information b is 0.3, with the of pending resource description information
Two degree of association are 0.5, and obtain key word Z and occur in resource description information b 6 times, key word Z and resource description information b
First degree of association is 0.5, is 0.2 with the second degree of association of the resource description information of pending resource;The most described accuracy determines
Device determines the degree of association=3*0.3*0.5=0.45 between key word Y and every other resource description information, key word Z and institute
There is the degree of association=6*0.5*0.2=0.6 between other resource description information;Further, described accuracy determines that device is by key word Y
And the degree of association between every other resource description information and the degree of association between key word Z and every other resource description information are entered
Row processes, and such as asks for both meansigma methodss, quadratic sum etc., and the result after processing is believed as described pending resource description
The degree of association between breath and every other resource description information.
It should be noted that the example above is only better described technical scheme, rather than the limit to the present invention
System, it should be appreciated by those skilled in the art that any according to described distributed intelligence and other relevant informations described, determine described in treat
Process the realization of the degree of association between resource description information and/or its each key word comprised and every other resource description information
Mode, should be included in the scope of the present invention.
Wherein, described accuracy determine device based on determined by the degree of association determine described pending resource description information
Overall accuracy and/or the implementation of the accuracy of each key word that comprised of described pending resource description information,
Step S4 the most in the embodiment shown in fig. 1 is described in detail, and is incorporated herein by reference, repeated no more.
One of preferably, the method according to the invention also include described accuracy determine device according to described in wait to locate
The accuracy information of reason resource description information and described resource thereof, set up or update the step of resource information bank.
Such as, described accuracy determines that key word X's that device determines that described pending resource description information comprises is accurate
Degree is 0.8, and the accuracy of key word Y is 0.1, and the most described accuracy determines that device is according to the accuracy of key word X and key word Y
Accuracy and described pending resource, set up or update resource information bank.
Preferably, described accuracy determines that device is by the link address information of described pending resource affiliated web site, described
The evaluation of estimate information etc. of pending resource is stored in described resource information bank.
According to the method for the present embodiment, multiple identical at other by the key word that pending resource description information is comprised
Or the distribution situation in the description information of similar resource and the analysis of other relevant informations, it is possible to more accurately determine pending
The degree of association between resource description information and/or its key word comprised and other resource description information, thus more precisely sentence
The accuracy of disconnected pending resource description information.
Fig. 4 show a preferred embodiment of the present invention according to determined by the accuracy information of resource description information
Resource is performed the flow chart of corresponding operating.
In step s 8, described accuracy determines that device obtains the behavior relevant information relevant to user behavior.Wherein, institute
State user behavior to include but not limited to: 1) user's initiative provide service behavior;Such as, user input query sequence is concurrent
Sending described search sequence etc., the most such as, user controls mouse makes cursor dwell to ask for the recommendation of this resource in a resource
Grade etc.;2) user triggers the behavior that resource information shows, such as, user opens a Webpage etc..Wherein, described behavior
Relevant information includes but not limited to: 1) the behavior operation information performed by user, such as, the behavioural information of request search, example again
As, the behavioural information etc. of request display resource recommendation grade;2) list entries that user is inputted, such as, user is inputted
List entries etc. for retrieval.
It should be noted that the example above is only better described technical scheme, rather than the limit to the present invention
System, it should be appreciated by those skilled in the art that any user behavior relevant to resource, should be included in the scope of the present invention.
Then, in step s 9, described accuracy determines that device determines pending money according to described behavior relevant information
Source.
Such as, described accuracy determines the list entries for retrieval that device inputs according to user, is obtained by after retrieving
Retrieval result in select pending resource, the mode of the pending resource of this selection includes but not limited to: randomly choose, based on
Number of clicks selects.The most such as, described accuracy determines the device position according to cursor dwell, corresponding to this position
Resource is as pending resource.The most such as, described accuracy determines that device opens a Webpage according to user, by this webpage
The resource comprised in the page is as pending resource etc..
Then, in step slo, described accuracy determines that device comes in described resource information according to described pending resource
Storehouse is inquired about, to obtain the accuracy information of resource description information corresponding to described pending resource.Wherein, described resource
Foundation and the renewal process of information bank are described in detail the most in the embodiment shown in fig. 3, and are incorporated herein by reference, no
Repeat again.
Then, in step s 11, described accuracy determines that device is believed according to the resource description that described pending resource is corresponding
The accuracy information of breath, performs operate corresponding with described user behavior.
Such as, for by retrieval result selects the described pending resource that obtains, described accuracy determine device according to
The accuracy information of the resource description information that described pending resource is corresponding, adjusts this pending resource in retrieval result
Sequence, and generate presenting information, so that described presenting information is supplied to described user according to the ranking results after adjusting.Example again
As, described accuracy determines that device position based on cursor dwell obtains pending resource, and the most described accuracy determines that device will
The accuracy information of the resource description information that the described pending resource that obtained is corresponding shows in the page at this cursor place,
Preferably, show in the way of contingent window and closing on this cursor position etc..The most such as, described accuracy determine device based on
The webpage that family is opened is to obtain pending resource, and the most described accuracy determines that device is by corresponding for the described pending resource obtained
The accuracy information of resource description information show in the web page.
It should be noted that the example above is only better described technical scheme, rather than the limit to the present invention
System, it should be appreciated by those skilled in the art that the accuracy letter of any resource description information corresponding according to described pending resource
Breath, performs the implementation that operate corresponding with described user behavior, should be included in the scope of the present invention.
According to the method for the present embodiment, can by determined by the accuracy of resource description information be applied to multiple occasion, example
As: 1) it is applied to searching system, so that the inaccurate resource of resource description information sorts rearward, make the sequence of retrieval result more
Rationally;2) be applied to commending system, such as, based on determined by the accuracy of resource description information recommend resource to user,
To improve the utilization rate of resource;3) prompt system, such as, based on determined by the accuracy of resource description information point out user
The description possible accuracy of this resource is relatively low etc..
Fig. 5 shows that the accuracy of the accuracy information for determining resource description information of one aspect of the invention determines
Device schematic diagram.Wherein, determine that device includes selecting device the 1, first acquisition device 2 and first true according to the accuracy of the present invention
Determine device 3.
Multiple description information that described selection device 1 is comprised by pre-established resource description information set select to treat
Reason resource description information, wherein, each resource description information in the plurality of resource description information is used to describe a money
Other resource description information arbitrary in source, and the resource described by each resource description information and this resource description information set
Described resource is similar or identical.Wherein, described resource includes but not limited to: 1) picture category resource;2) audio class resource;3)
Video class resource;4) program bag class resource etc..
Wherein, the mode of pre-established resource description information set includes but not limited to:
1) artificial next pre-established resource description information set.
A) for picture category resource, operator when setting up resource description information set, view-based access control model effect judges
Multiple resources are the most same or similar.Such as, if for picture category resource resource A1 and resource B1 phase in visual effect
With, being only merely and there are differences at aspects such as background color, size, regional areas, then operator judge resource A1 and resource
B1 is similar.
B) for video class resource, operator, when setting up resource description information set, judge based on resource plot
Multiple resources are the most same or similar.Such as, if resource A2 is identical with the main action of resource B2, simply in image resolution
The aspect such as rate, compressed format is different, then operator judge that resource A2 is similar to resource B2.
C) for audio class resource, operator, when setting up resource description information set, judge based on auditory effect
Multiple resources are the most same or similar.Such as, resource A3 and resource B3 are identical on auditory effect, be different only in that resource A3 with
The aspects such as the lyrics of resource B3, compressed format are different, then operator judge that resource A3 is similar to resource B3.
D) for program bag class resource, based on program source code, operator judge that multiple resource is the most identical or phase
Seemingly.Such as, resource A4 and the source code of resource B4 are simply in the name of variable, pointer, array etc. or to program source code
There is difference in the aspects such as explanation, then operator judge that resource A4 is similar to resource B4.
2) pre-established resource description information set is carried out based on resource cluster.This sets up mode will in the embodiment shown in fig. 6
Described in detail.
It should be noted that the example above is only better described technical scheme, rather than the limit to the present invention
System, it should be appreciated by those skilled in the art that and any determine the most same or analogous mode of resource, and based on same or similar
Resource carrys out the implementation of pre-established resource description information set, should be included in the scope of the present invention.
Multiple description information that selection device 1 is comprised by pre-established resource description information set select pending money
The selection mode of Source Description information includes but not limited to:
1) multiple description information that described selection device 1 is comprised by pre-established resource description information set at random are selected
Select pending resource description information.
2) described selection device 1 includes identifying device (not shown), and described identification device is according to the plurality of resource description
The network related information that resource described by information is corresponding, is identified the plurality of resource description information, will identify institute
The user obtained generates resource description information as described pending resource description information.
Such as, the plurality of resource description information includes the resource description information of resource A from website A ', from website
The resource description information of resource B of B ' and the resource description information of resource C from website C ', described identification device is according to predetermined
Authoritative website list determine that described website A ' and website B ' are authoritative website, website C ' is inauthoritativeness website, therefore, described knowledge
Other device according to the resource description information of resource A, the resource description information of resource B and the resource description information institute of resource C from
Website, identify the resource description information of resource C from inauthoritativeness website, and using the resource description information of resource C as described
Pending resource description information.
Preferably, described network related information include following at least one:
A) link address information of the resource that this network related information is corresponding.Specifically, described identification device is according to resource
Link address information in the pre-determined text information that comprises, such as: i) bbs;ii)blog;Iii) SNS etc., identify this resource
Corresponding resource description information is that user generates resource description information, and then using resource description information corresponding for this resource as treating
Process resource description information.Such as, the resource described by the plurality of resource description information includes resource A and resource B.Wherein,
The link address information of resource A is " www.222.com ", and the link address information of resource B is " bbs.444.com ", then described
Identify that device comprises " bbs " according to the link address information of resource B, identify that the resource description information of resource B is that user generates money
Source Description information, and using the resource description information of resource B as pending resource description information.
B) the page feature information of webpage belonging to the resource that this network related information is corresponding.Specifically, described identification device
According to the code of webpage belonging to resource being analyzed the page feature information of gained, such as, model category feature information, be contained in
The particular text information etc. such as such as " blog " in page subject matter, " album ", determine the resource belonging to this webpage
Resource description information be that user generates resource description information, and then using resource description information corresponding for this resource as pending
Resource description information.Preferably, described model category feature information includes;1) the model class text such as " main building ", " 1st floor ", " building-owner "
Information;2) the model class formation information of the display module etc. that multiple stacking shows and structure is identical is comprised.
C) the page feature information of the webpage that the resource affiliated web site that this network related information is corresponding is comprised.Specifically,
Described identification device is analyzed the Webpage of this website of gained according to the web page code being comprised resource affiliated web site
Characteristic information, such as, occurs in the particular texts such as such as " blog " in the page subject matter of multiple webpage, " home videos "
Information, the model class formation information etc. occurred in multiple webpage, determine the resource description information of the resource belonging to this webpage
Resource description information is generated for user, and then using resource description information corresponding for this resource as pending resource description information.
It is highly preferred that described identification device is according at least one in above-mentioned three network related information, come the plurality of
Resource description information is identified, so that the user identifying gained is generated resource description information as described pending resource description
Information.
Such as, as pre-determined text information " bbs " comprised during described identification device obtains the link address information of resource,
Analyze whether the page feature information of webpage belonging to resource comprises model category feature information further, and when page feature information bag
During containing model category feature information, this resource identification is that user generates resource description information by, and this resource is treated as described
Process resource description information.
It should be noted that the example above is only better described technical scheme, rather than the limit to the present invention
System, it should be appreciated by those skilled in the art that any can be for identifying that resource generates the net of resource description information to obtain user
Network relevant information, should be included in the scope of the present invention.
Then, described first acquisition device 2 obtains each key word that described pending resource description information comprised and exists
Distributed intelligence in other resource description information described.Wherein, one or many is only comprised when described pending resource description information
During the key word of individual separation, described first acquisition device 2 directly obtains the one or more key word in other resources described
Distributed intelligence in description information;When described pending resource description information comprises one or more text, described first obtains
Fetching put 2 the one or more text is cut word, duplicate removal etc. process obtain described pending resource description information bag
The key word contained.
Wherein, described distributed intelligence include following at least one:
1) total degree that each key word described occurs in described every other resource description information.Such as, pre-established
Resource description information set include describe resource A pending resource description information a, describe resource B resource description letter
Breath b and resource description information c of description resource C, described first acquisition device 2 obtains the key word that resource description information a comprises
Including key word a1 and key word a2 and obtain key word a1 and key word a2 and occur in resource description information b 2 times, in money
Occurring 1 time in Source Description information c, the most described first acquisition device 2 obtains the key word a1 that pending resource description information a comprises
The total degree occurred in described resource description information b with resource description information c with key word a2 is 2+1=3 time.
2) number of times that each key word described occurs in described every other resource description information respectively.Such as, build in advance
Vertical resource description information set includes describing pending resource description information d of resource D, describing the resource description of resource E
Information e and resource description information f describing resource F;Described first acquisition device 2 obtains pending resource description information d and comprises
Key word include key word d1 and key word d2 and obtain key word d1 in resource description information e and resource description information f
Occur 5 times, obtain key word d2 and occur 3 times in resource description information e with resource description information f.
3) identification information of other resource description information described of arbitrary key word in each key word described is comprised.Example
As, pre-established resource description information set includes describing pending resource description information g of resource G, describing the money of resource H
Source Description information h and resource description information i describing resource I;Described first acquisition device 2 obtains pending resource description information
The key word that g comprises includes key word g1 and determines that resource description information h comprises key word g1, does not wraps in resource description information i
Containing key word g1, the identification information that the most described first acquisition device 2 obtains the resource description information comprising key word g1 is h.
4) quantity of other resource description information described of at least one key word described is comprised.Such as, pre-established money
Source Description information aggregate include describe resource J pending resource description information j, describe resource K resource description information k with
Describing the resource description information 1 of resource L, described first acquisition device 2 obtains the key word that pending resource description information j comprises
Including key word j1 and key word j2 and determine resource description information k comprises key word j1, resource description information 1 comprises
Key word j2, the most described first acquisition device 2 obtains and comprises other moneys described of at least one in key word j1 and key word j2
The quantity of Source Description information is 2.
5) quantity of other resource description information described comprising at least one key word described accounts for described all resources and retouches
State the ratio of the quantity of information.
6) quantity of other resource description information that each key word in each key word described is occurred accounts for all money
The ratio of the quantity of Source Description information.Such as, a key word occurs in 4 other resource description information, and all resources
The quantity of description information is 10, then this key word accounts for the quantity of all resource description information in the quantity of other resource description information
Ratio be 0.4.
Then, described first determine device 3 according to described distributed intelligence, determine described pending resource description information and/
Or the degree of association between its each key word comprised and every other resource description information, to obtain this pending resource description letter
The accuracy information of breath.
Wherein, described first determine device 3 according to described distributed intelligence, determine described pending resource description information and/
Or the mode of the degree of association between its each key word comprised and every other resource description information includes but not limited to:
1) directly using distributed intelligence as described pending resource description information and/or its each key word comprised and institute
There is the degree of association between other resource description information.Such as, described first acquisition device 2 obtains and comprises described pending resource extremely
The ratio of the quantity that the quantity of other resource description information described of a few key word accounts for described all resource description information is
0.8, the most described first determines that device 3 determines the pass between described pending resource description information and every other resource description information
Connection degree is 0.8.The most such as, the quantity of other resource description information that each key word in each key word described is occurred accounts for
The ratio of the quantity of all resource description information is 0.4, and the most described first determines that device 3 determines this key word and every other money
The degree of association between Source Description information is 0.4.
2) carry out distributed intelligence processing obtained result as described pending resource description information and/or its
The degree of association between each key word comprised and every other resource description information.Specifically, carry out distributed intelligence processing
Mode includes: a) obtain the described degree of association according in distributed intelligence, such as: i) distributed intelligence entered with predetermined threshold
Row compares, and determines described pending resource description information and/or its each key word comprised and institute according to comparative result
There is the relevance level between other resource description information;Ii) ask for distributed intelligence to retouch with the resource in resource description information set
State the ratio of information sum, and determine described pending resource description information and/or its each pass comprised according to gained ratio
The degree of association between keyword and every other resource description information;B) described pending money is obtained according to multinomial in distributed intelligence
The degree of association between Source Description information and/or its each key word comprised and every other resource description information, such as: i) by two
Item distributed intelligence be used for described pending resource description information and/or its each key word comprised and every other money
The degree of association between Source Description information;Ii) multinomial distribution information is normalized, and the value of normalized gained is entered
Row sue for peace, average, ask logarithm and etc. process, using the value of gained as described pending resource description information and/or its
The degree of association between each key word comprised and every other resource description information;Iii) come multinomial distribution according to predetermined formula
Information carries out calculation process, and using the value of calculation process gained as described pending resource description information and/or its comprise
The degree of association etc. between each key word and every other resource description information.
Such as, described first acquisition device 2 obtains each key word of pending resource and retouches in described every other resource
Stating the total degree occurred in information is 10 times, and the most described first determines that device 3 is higher than the first predetermined threshold based on this total degree, comes
Determine that the degree of association between described pending resource description information and every other resource description information is senior.
The most such as, described first acquisition device 2 obtains key word that pending resource description information comprises described all
The number of times occurred in other resource description information is 5 times, and the most described first determines that device 3 is less than the second predetermined threshold based on this number of times
Value, determines that the degree of association between described pending resource description information and every other resource description information is rudimentary.
The most such as, described first acquisition device 2 obtains and comprises key word Y's that described pending resource description information comprises
The quantity of other resource description information described is 6, and obtains key word X, key that described pending resource description information comprises
The total degree that word Y and key word Z occurs in described every other resource description information is 60 times, and the most described first determines device
3 retouch comprising the quantity of other resource description information described in this key word Y in described every other resource with each key word
State the ratio 6/60=0.1 of the total degree occurred in information as the pass between this key word Y and every other resource description information
Connection degree.
The most such as, described first acquisition device 2 obtains each key word of pending resource in described every other resource
The total degree occurred in description information is 20 times, and based on comprising other moneys described in arbitrary key word in each key word described
The identification information of Source Description information obtains and comprises other resource description information described of arbitrary key word in each key word described
Quantity be 5, the most described first determines that each key word of described pending resource is retouched by device 3 in described every other resource
That states the total degree and the acquisition that occur in information comprises other resource descriptions described of arbitrary key word in each key word described
The ratio of number 20/5=4 of information is as described pending resource description information and associating between every other resource description information
Degree.
The most such as, described first acquisition device 2 obtains each key word of pending resource in described every other resource
The total degree occurred in description information is 10 times, and in described resource description information set, all of key word quantity is 50, comprises
The quantity of other resource description information described of described arbitrary key word accounts for the ratio of the quantity of described all resource description information
Be 0.5, then first determines that device 3 is according to the first predetermined formula: described pending resource description information is retouched with every other resource
That states that each key word of the degree of association between information=pending resource occurs in described every other resource description information is total
Other resource descriptions described of all of key word quantity+described arbitrary key word in number of times/described resource description information set
The quantity of information accounts for the ratio of the quantity of described all resource description information, determines described pending resource description information and institute
There is the degree of association=10/50+0.5=0.7 between other resource description information.
The most such as, described first acquisition device 2 obtains key word V and key word W that pending resource comprises in described institute
Having the total degree occurred in other resource description information is 10 times, it is thus achieved that the key word V that pending resource comprises is described all
The number of times occurred in other resource description information is 3 times, comprises at least one key word in described key word V and key word W
The ratio of the quantity that the quantity of other resource description information described accounts for described all resource description information is 0.9, then first determines
Device 3 is according to the second predetermined formula: the key word that described pending resource description information comprises is believed with every other resource description
Number of times/pending resource that the degree of association between breath=this key word occurs in described every other resource description information comprises
The total degree * that each key word occurs in described every other resource description information comprises described at least one key word described
The quantity of other resource description information accounts for the ratio of the quantity of described all resource description information, determines that key word V is with all
The degree of association=3/10*0.9=0.27 between other resource description information.
It should be noted that the example above is only better described technical scheme, rather than the limit to the present invention
System, it should be appreciated by those skilled in the art that any according to described distributed intelligence, determine described pending resource description information and/
Or the implementation of the degree of association between its each key word comprised and every other resource description information, should be included in this
In bright scope.
Wherein, described accuracy information include following at least one: 1) described pending resource description information overall accurate
Exactness;2) accuracy of each key word that described pending resource description information is comprised.
Specifically, described first determine device 3 based on determined by pending resource description information and every other resource
The mode of the overall accuracy that the degree of association between description information obtains pending resource description information includes but not limited to: 1) straight
Connect the degree of association between described pending resource description information and every other resource description information as pending resource description
The overall accuracy of information.2) degree of association between described pending resource description information and every other resource description information is entered
The result that row process is obtained is as the overall accuracy of pending resource description information.Such as, by described pending money
The degree of association between Source Description information and every other resource description information is retouched as pending resource with the product of predefined weight value
State the overall accuracy of information.The most such as, by between described pending resource description information and every other resource description information
Each key word that the degree of association is asked for square or the result of 3 powers is comprised as described pending resource description information accurate
Degree etc..3) degree of association between each key word described pending resource comprised and every other resource description information is asked
The result obtained with summation after, weighted sum, quadrature, normalization etc. as pending resource description information overall accurately
Degree.
Described first determines that each key word that device 3 is comprised based on described pending resource is retouched with every other resource
State the mode of the degree of association between the information accuracy of each key word to determine described pending resource description information and comprised
Include but not limited to;1) directly each key word that described pending resource description information is comprised is retouched with every other resource
State the accuracy of each key word that each degree of association between information is comprised respectively as described pending resource description information.
2) degree of association between each key word described pending resource description information comprised and every other resource description information
Carry out processing each key word that each result obtained is comprised respectively as described pending resource description information
Accuracy.Such as, each between each key word described pending resource comprised and every other resource description information
Each that the individual degree of association is asked for square respectively or each results of 3 powers is comprised as described pending resource description information closes
The accuracy of keyword.The most such as, each key word described pending resource comprised and every other resource description information
Between the product of each degree of association and predefined weight comprised respectively as described pending resource description information each is crucial
The accuracy of word.
It should be noted that the example above is only better described technical scheme, rather than the limit to the present invention
System, it should be appreciated by those skilled in the art that any based on determined by the degree of association obtain the total of pending resource description information
The implementation of the accuracy of each key word that body accuracy and/or described pending resource description information are comprised, all should
Within the scope of the present invention.
Accuracy according to the present invention determines that device can be by the key comprising the resource description information of a resource
Word distribution situation in the resource description information of other multiple same or similar resources, determine this resource description information or its
The key word comprised and the degree of association of other resource description information, the resource described by this resource description information and other money
Resource described by Source Description information is same or similar, therefore, this degree of association can reflect this resource description information or its comprise
The description accuracy of key word.Accuracy according to the present invention determines that device is particularly suited for determining the money that user generates resource
The description accuracy of Source Description information.
One of preferred version as the present invention, Fig. 6 show a preferred embodiment of the invention based on resource cluster
The accuracy carrying out pre-established resource description information set determines device schematic diagram.Accuracy according to the present embodiment determines device bag
Include the 3rd acquisition device 4, clustering apparatus 5 and construction device 6.
Described 3rd acquisition device 4 obtains multiple resource.Wherein, described 3rd acquisition device 4 obtains the side of multiple resource
Formula includes but not limited to: 1) by obtaining the plurality of resource in multiple websites;2) described many by the resources bank of pre-stored obtains
Individual resource etc..
It should be noted that the example above is only better described technical scheme, rather than the limit to the present invention
System, it should be appreciated by those skilled in the art that the implementation of the multiple resource of any acquisition, should be included in the scope of the present invention
In.
Then, the information that described clustering apparatus 5 self is comprised according to the plurality of resource, the plurality of resource is carried out
Cluster, to obtain one or more groups cluster resource, wherein, often group cluster resource includes one or more same or analogous money
Source.Wherein, described clustering apparatus 5 uses corresponding cluster mode according to resource type.Such as, for picture category resource, institute
State picture element information that clustering apparatus 5 comprises according to picture, the color histogram information of picture, local invariant feature (SIFT,
Scale-invariant feature transform), textural characteristics (HTD, Homogeneous Texture
Descriptor), color characteristic (SCD) etc., carry out picture cluster.The most such as, for video class resource, described clustering apparatus
5 cluster according to the size of video resource, form, the information such as sectional drawing of identical time point.The most such as, audio class is provided
Source, described clustering apparatus 5 according to the form of audio frequency, size, the average pitch of audio resource, audio resource on each time point
The information such as tone cluster.The most such as, for program bag class resource, described clustering apparatus 5 comprises according to program bag
Source code information etc. clusters.
It should be noted that the example above is only better described technical scheme, rather than the limit to the present invention
System, it should be appreciated by those skilled in the art that and any resource clusters to obtain one or more groups cluster resource, wherein, often
Group cluster resource includes the cluster mode of one or more same or analogous resource, should be included in the scope of the present invention.
Then, described construction device 6 often organizes, according to described, the resource description information that same or analogous resource is corresponding, builds
Vertical described resource description information set.
Such as, described clustering apparatus 5 obtains one group of cluster resource A1, one group of cluster resource A2, one group of cluster resource A3, institute
State construction device 6 according to resource description corresponding to resource description information corresponding to cluster resource a1 that comprises of resource A1, resource a2
Information and resource description information corresponding to resource a3, set up described resource description information set.
Preferably, described construction device 6 is additionally based upon cluster resource A2 or resource corresponding to the cluster resource that comprises of resource A3
Description information, sets up another resource description information set.
Fig. 7 shows the accurate of the accuracy information for determining resource description information of a preferred embodiment of the invention
Degree determines device schematic diagram.Accuracy according to the present embodiment determines that device includes selecting device the 1, first acquisition device 2, first
Determine device 3 and the second acquisition device 7;Described first determines that device 3 also includes that son determines device 301.Wherein, device 1 is selected
And first acquisition device 2 described in detail the most with reference to the embodiment shown in FIG. 5, and be incorporated herein by reference, the most superfluous
State.
Described second acquisition device 7 obtains other relevant informations for determining the described degree of association.
Wherein, other relevant informations described include following at least one;
1) resource described by other resource description information described of arbitrary key word in each key word described is comprised
Authoritative.
Wherein, the authoritative mode that described second acquisition device 7 obtains resource includes but not limited to: a) obtain pre-stored
The authority of this resource;B) characteristic information based on this resource affiliated web site determines the authority of this resource.Such as, described
Whether the second acquisition device 7 visit capacity based on this website, this website are included in predetermined authoritative website, the list of material website
In, the quantity of the resource from this website that comprises in information bank whether exceed predetermined threshold and information bank comprise from this
Whether the quality information of the resource of website is high-quality etc., determines the authority of this resource.
2) each key word in described all key words and comprise this key word each other resource description information between
The first degree of association.
Wherein, described second acquisition device 7 obtains key word and each other resource description information comprising this key word
Between the mode of the first degree of association include but not limited to:
A) key word obtaining pre-stored is relevant to first between each other resource description information comprising this key word
Degree;Such as, other resource description information including key word X are resource description information b and resource description information c, and described
The first degree of association between pre-stored key word X and resource description information b in the storage device that second acquisition device 7 can be accessed by
Being 2, the first degree of association between key word X and resource description information c is 3, then the second acquisition device 7 obtains the key word of pre-stored
X and comprise this key word other resource description information b and c between the first degree of association be respectively 2 and 3.
B) described second acquisition device 7 based on following at least one determine key word with comprise this key word one its
Described first degree of association between his resource description information, with determine this key word respectively with comprise this key word each other money
The first degree of association between Source Description information:
I) number of times that this key word occurs in other resource description information;Such as, the second acquisition device 7 is by this pass
The key word sum that the number of times that keyword occurs in other resource description information and these other resource description information are comprised
Ratio, as the first degree of association between this key word and this other resource description information.
Ii) text type of the text message at this key word place;Wherein, described text message is contained in other resources and retouches
State in information, and described text type include but not limited to: title class text, Anchor Text class text, at webpage belonging to this resource
In the context class text etc. adjacent with resource;Such as, when the text type comprising this key word is title class text, then described
Second acquisition device 7 determines that the first degree of association of this key word is senior.
Iii) number of times that this key word occurs in each text type that other resource description information comprise respectively
And the predefined weight value of each text type;Such as, described second acquisition device 7 obtains this key word at these other resource descriptions
The title class text that information comprises occurs 1 time occur 8 times in context class text, and the predefined weight value of title class text
Being 0.6, the predefined weight value of context class text is 0.3, then the second acquisition device 7 determines the predefined weight of different text type
Value and this key word occur in the sum of products=0.6*1+0.3*8=3 of the number of times in dissimilar text, and are made by this sum of products
For this key word and comprise this key word these other resource description information between the first degree of association.
It should be noted that the example above is only better described technical scheme, rather than the limit to the present invention
System, it should be appreciated by those skilled in the art that any these other resource descriptions for determining this key word with comprise this key word
The implementation of the first degree of association between information, such as, the number of times that key word is occurred in other resource description information
It is multiplied by the meansigma methods of the predefined weight value of each text type of the text message at this key word place, obtains described first phase
Guan Du etc., should be included in the scope of the present invention.
3) each key word in described all key words and the second degree of association between described pending resource description information.
Wherein, each key word that described second acquisition device 7 obtains in described all key words is believed with described pending resource description
The acquisition mode of the second degree of association between breath, obtains each key in described all key words with described second acquisition device 7
Word and comprise this key word other resource description information between the acquisition mode of the first degree of association same or similar, and to quote
Mode be incorporated herein, repeat no more.
It should be noted that the example above is only better described technical scheme, rather than the limit to the present invention
System, it should be appreciated by those skilled in the art that any other relevant informations for determining the described degree of association and any acquisition are used
In the implementation of other relevant informations determining the described degree of association, should be included in the scope of the present invention.
Need it is further noted that the first acquisition device 2 obtain that described pending resource description information comprised each
The operation of individual key word distributed intelligence in other resource description information described obtains with the second acquisition device 7 and is used for determining institute
The operation of other relevant informations stating the degree of association there is no sequencing.
Then, described son determine device 301 according to described distributed intelligence and other relevant informations described, determine described in treat
Process the degree of association between resource description information and every other resource description information, to obtain this pending resource description information
Accuracy.Wherein, described son determine device 301 according to described distributed intelligence and other relevant informations described, determine described in treat
The mode processing the degree of association between resource description information and every other resource description information includes but not limited to:
1) described son determines that device 301 first determines other resources comprising at least one key word based on described distributed intelligence
Description information, further according to determined by comprise at least one key word every other resource description information and described other be correlated with
Information determines described pending resource description information and/or its each key word comprised and every other resource description information
Between the degree of association.Such as, described son determines that device 301 is first based on comprising the institute of at least one key word in each key word described
State the identification information of other resource description information, determine other moneys described comprising at least one key word in resource description set
Source Description information includes resource description information a describing resource A, and then, described son determines the device 301 power further according to resource A
Prestige is senior, determines that the degree of association between described pending resource description information and every other resource description information is senior.
The most such as, described son determines that device 301 is first based on comprising the institute of at least one key word in each key word described
State the identification information of other resource description information, determine other moneys described comprising at least one key word in resource description set
Source Description information includes resource description information b describing resource B and describes resource description information c of resource C, and determines description money
Resource description information b of source B comprises key word Y, and resource description information c describing resource C comprises key word X and key word Y, institute
State son and determine that device 301 the first degree of association based on key word X Yu resource description information c is 0.6, determine this key word X with
The degree of association between every other resource description information is 0.6, and the first degree of association based on key word Y Yu resource description information b
It is 0.8 and the first degree of association of key word Y and resource description information c is 0.4, determines this key word Y and every other resource
The degree of association=0.8+0.4=1.2 between description information.
2) son determines that device 301 is according at least at least one in distributed intelligence and other relevant informations described
Item determines the described degree of association.Specifically, described son determines that device 301 adjusts described distribution based on other relevant informations described
The value that information is comprised, and based on adjust after result determine described pending resource description information and/or its comprise each
The degree of association between individual key word and every other resource description information.
Such as, described son determines that device 301 obtains key word X resource description information a in resource description information set
Middle appearance 2 times, and the first degree of association obtaining this key word X and resource description information a is 0.6, the most described son determines device 301
With this first degree of association as Dynamic gene, determine that the described degree of association between this key word X and every other resource description information is
0.6*2=1.2.
The most such as, described son determines that device 301 obtains key word Y resource description information in resource description information set
Occurring 3 times in b, key word Y is 0.3 with the first degree of association of resource description information b, with the second of pending resource description information
Degree of association is 0.5, and obtains key word Z and occur in resource description information b 6 times, the of key word Z and resource description information b
One degree of association is 0.5, is 0.2 with the second degree of association of the resource description information of pending resource;The most described son determines device 301
Determine the degree of association=3*0.3*0.5=0.45 between key word Y and every other resource description information, key word Z with all its
The degree of association=6*0.5*0.2=0.6 between his resource description information;Further, described son determines that device 301 is by key word Y and institute
Have at the degree of association between other resource description information and the degree of association between key word Z and every other resource description information
Reason, such as asks for both meansigma methodss, quadratic sum etc., and the result after processing as described pending resource description information with
The degree of association between every other resource description information.
It should be noted that the example above is only better described technical scheme, rather than the limit to the present invention
System, it should be appreciated by those skilled in the art that any according to described distributed intelligence and other relevant informations described, determine described in treat
Process the realization of the degree of association between resource description information and/or its each key word comprised and every other resource description information
Mode, should be included in the scope of the present invention.
Wherein, described son determine device 301 based on determined by the degree of association determine described pending resource description information
Overall accuracy and/or the implementation of the accuracy of each key word that comprised of described pending resource description information,
With described first in the embodiment shown in Fig. 5 determine device 3 based on determined by described pending resource description information and/or
The degree of association between its each key word comprised and every other resource description information, obtains this pending resource description information
The implementation of accuracy information same or similar, be incorporated herein by reference, repeat no more.
One of preferred version as the present invention, described accuracy determines that device also includes updating device (not shown).Institute
State updating device according to the accuracy information of described pending resource description information and described resource thereof, set up or update money
The step in source information storehouse.
Such as, the described first accuracy determining key word X that device 3 determines that described pending resource description information comprises
Being 0.8, the accuracy of key word Y is 0.1, and the most described updating device is according to the accuracy of key word X and the accuracy of key word Y
And described pending resource, set up or update resource information bank.
Preferably, described updating device is by the link address information of described pending resource affiliated web site, described pending
The evaluation of estimate information etc. of resource is stored in described resource information bank.
Accuracy according to the present embodiment determines device, by the key word that comprises pending resource description information at it
Distribution situation in the description information of his multiple same or similar resources and the analysis of other relevant informations, it is possible to more precisely
Determine the degree of association between pending resource description information and/or its key word comprised and other resource description information, thus more
Adequately judge the accuracy of pending resource description information.
Fig. 8 show a preferred embodiment of the present invention according to determined by the accuracy information of resource description information
The accuracy that resource performs corresponding operating determines device schematic diagram.Accuracy according to the present embodiment determines that device includes
Four acquisition device 8, second determine device 9, inquiry unit 10 and perform device 11.
Described 4th acquisition device 8 obtains the behavior relevant information relevant to user behavior.Wherein, described user behavior bag
Include but be not limited to: 1) user's initiative provide service behavior;Such as, user input query sequence send described inquiry sequence
Row etc., the most such as, user controls mouse makes cursor dwell to ask for the recommendation grade etc. of this resource in a resource;2) user
Triggering the behavior that resource information shows, such as, user opens a Webpage etc..Wherein, described behavior relevant information includes
But it is not limited to: 1) behavior operation information performed by user, such as, the behavioural information of request search, the most such as, request display money
The behavioural information etc. of source recommendation grade;2) list entries that user is inputted, such as, the input for retrieval that user is inputted
Sequence etc..
It should be noted that the example above is only better described technical scheme, rather than the limit to the present invention
System, it should be appreciated by those skilled in the art that any user behavior relevant to resource, should be included in the scope of the present invention.
Then, described second determines that device 9 determines pending resource according to described behavior relevant information.
Such as, described second determines the list entries for retrieval that device 9 inputs according to user, is obtained by after retrieving
Retrieval result in select pending resource, the mode of the pending resource of this selection includes but not limited to: randomly choose, based on point
Hit number of times to select.The most such as, described second determines the device 9 position according to cursor dwell, by the money corresponding to this position
Source is as pending resource.The most such as, described second determines that device 9 opens a Webpage according to user, by this webpage page
The resource comprised in face is as pending resource etc..
Then, described inquiry unit 10 is inquired about in described resource information bank according to described pending resource, with
Obtain the accuracy information of resource description information corresponding to described pending resource.Wherein, described resource information bank foundation and
Renewal process is described in detail the most in the embodiment shown in fig. 7, and is incorporated herein by reference, repeats no more.
Then, described execution device 11 according to the accuracy information of resource description information corresponding to described pending resource,
Perform operate corresponding with described user behavior.
Such as, for by selecting the described pending resource that obtains in retrieval result, described execution device 11 is according to inquiry
The accuracy information of the resource description information that described pending resource that device 10 is obtained is corresponding, adjusts this pending resource
Sequence in retrieval result, and generate presenting information, to be provided by described presenting information according to the ranking results after adjusting
To described user.The most such as, described second determine device 9 position based on cursor dwell obtain pending resource, then described in hold
Luggage is put the accuracy information of resource description information corresponding to the 11 described pending resources obtained by inquiry unit 10 and is shown
In the page at this cursor place, it is preferred that show in the way of contingent window and closing on this cursor position etc..The most such as, institute
Stating second and determine that device 9 obtains pending resource based on the webpage that user opens, the most described execution device 11 is by inquiry unit
The accuracy information of the resource description information that the 10 described pending resources obtained are corresponding shows in the web page.
It should be noted that the example above is only better described technical scheme, rather than the limit to the present invention
System, it should be appreciated by those skilled in the art that the accuracy letter of any resource description information corresponding according to described pending resource
Breath, performs the implementation that operate corresponding with described user behavior, should be included in the scope of the present invention.
Accuracy according to the present embodiment determines device, can by determined by the accuracy of resource description information be applied to many
Plant occasion, such as: 1) it is applied to searching system, so that the inaccurate resource of resource description information sorts rearward, make retrieval result
Sequence the most reasonable;2) be applied to commending system, such as, based on determined by resource description information accuracy come to user
Recommend resource, to improve the utilization rate of resource;3) prompt system, such as, based on determined by the accuracy of resource description information
Point out the description possible accuracy of this resource of user relatively low etc..
It is obvious to a person skilled in the art that the invention is not restricted to the details of above-mentioned one exemplary embodiment, Er Qie
In the case of the spirit or essential attributes of the present invention, it is possible to realize the present invention in other specific forms.Therefore, no matter
From the point of view of which point, all should regard embodiment as exemplary, and be nonrestrictive, the scope of the present invention is by appended power
Profit requires rather than described above limits, it is intended that all by fall in the implication of equivalency and scope of claim
Change is included in the present invention.Should not be considered as limiting involved claim by any reference in claim.This
Outward, it is clear that " including ", a word was not excluded for other unit or step, and odd number is not excluded for plural number.In system claims, statement is multiple
Unit or device can also be realized by software or hardware by a unit or device.The first, the second word such as grade is used for table
Show title, and be not offered as any specific order.
Claims (23)
1. a method for the computer implemented accuracy information for determining resource description information, wherein, the method includes
Following steps:
Multiple resource description information that a is comprised by pre-established resource description information set select pending resource description to believe
Breath, wherein, each resource description information in the plurality of resource description information is used to describe a resource, and each resource
Resource described by description information and the resource described by other resource description information arbitrary in this resource description information set
Similar or identical;
B obtains each key word dividing in other resource description information described that described pending resource description information is comprised
Cloth information;
C, according to described distributed intelligence, determines that described pending resource description information and/or its each key word comprised are with all
The degree of association between other resource description information, to obtain the accuracy information of this pending resource description information.
Method the most according to claim 1, wherein, described distributed intelligence include following at least one:
The total degree that-each key word described occurs in described every other resource description information;
The number of times that-each key word described occurs in described every other resource description information respectively;
-comprise the identification information of other resource description information described of at least one key word in each key word described;
-comprise the quantity of other resource description information described of at least one key word described;
-the quantity of other resource description information described that comprises at least one key word described accounts for described all resource description information
The ratio of quantity;
The quantity of other resource description information that each key word in-each key word described is occurred accounts for all resource descriptions
The ratio of the quantity of information.
Method the most according to claim 1 and 2, wherein, the method is further comprising the steps of:
-obtain other relevant informations for determining the described degree of association;
Wherein, described step c is further comprising the steps of:
-according to described distributed intelligence and other relevant informations described, determine described pending resource description information and/or its bag
The degree of association between each key word contained and every other resource description information, to obtain the standard of this pending resource description information
Exactness information.
Method the most according to claim 3, wherein, other relevant informations described include following at least one:
-comprise the authority of the resource described by other resource description information described of arbitrary key word in each key word described
Property;
Each key word in-described all key words and comprise this key word each other resource description information between first
Degree of association;
Each key word in-described all key words and the second degree of association between described pending resource description information.
Method the most according to claim 1, wherein, described step a is further comprising the steps of:
-according to network related information corresponding to the resource described by the plurality of resource description information, the plurality of resource is retouched
The information of stating is identified, so that the user identifying gained is generated resource description information as described pending resource description information.
Method the most according to claim 5, wherein, described network related information include following at least one:
The link address information of the resource that-this network related information is corresponding;
The page feature information of webpage belonging to the resource that-this network related information is corresponding;
The page feature information of the webpage that resource affiliated web site corresponding to-this network related information is comprised.
Method the most according to claim 1, wherein, the method is further comprising the steps of:
-obtain multiple resource;
-the information that self comprised according to the plurality of resource, clusters the plurality of resource, to obtain one or more groups
Cluster resource, wherein, often group cluster resource includes one or more same or analogous resource;
Wherein, the method is further comprising the steps of:
-often organize, according to described, the resource description information that same or analogous resource is corresponding, set up described resource description information collection
Close.
Method the most according to claim 1, wherein, the method is further comprising the steps of:
-believe according to accuracy information and described resource, foundation or the more new resources of described pending resource description information
Breath storehouse.
Method the most according to claim 8, wherein, the method is further comprising the steps of:
-obtain the behavior relevant information relevant to user behavior;
-according to described behavior relevant information, determine pending resource;
-inquire about in described resource information bank according to described pending resource, corresponding to obtain described pending resource
The accuracy information of resource description information;
-according to the accuracy information of resource description information corresponding to described pending resource, perform corresponding to described user behavior
Operation.
Method the most according to claim 9, wherein, described user behavior information include following at least one:
The type of-user operation;
The object of-user operation;
The input content inputted in-user input operation.
11. methods according to claim 1, wherein, described accuracy information include following at least one:
The overall accuracy of-described pending resource description information;
The accuracy of each key word that-described pending resource description information is comprised.
The accuracy of 12. 1 kinds of computer implemented accuracy information for determining description information determines device, wherein, and this standard
Exactness determines that device includes:
Select device, select to treat in the multiple resource description information comprised by pre-established resource description information set
Reason resource description information, wherein, each resource description information in the plurality of resource description information is used to describe a money
Other resource description information arbitrary in source, and the resource described by each resource description information and this resource description information set
Described resource is similar or identical;
First acquisition device, each key word comprised for obtaining described pending resource description information provide at described other
Distributed intelligence in Source Description information;
First determine device, according to described distributed intelligence, determine described pending resource description information and/or its comprise each
The degree of association between key word and every other resource description information, to obtain the accuracy letter of this pending resource description information
Breath.
13. accuracy according to claim 12 determine device, wherein, described distributed intelligence include following at least one:
The total degree that-each key word described occurs in described every other resource description information;
The number of times that-each key word described occurs in described every other resource description information respectively;
-comprise the identification information of other resource description information described of at least one key word in each key word described;
-comprise the quantity of other resource description information described of at least one key word described;
-the quantity of other resource description information described that comprises at least one key word described accounts for described all resource description information
The ratio of quantity;
The number of times that each key word in-each key word described occurs in described every other resource description information accounts for all
The ratio of the quantity of resource description information.
14. determine device according to the accuracy described in claim 12 or 13, and wherein, this accuracy determines that device also includes:
Second acquisition device, for obtaining for determining other relevant informations of the described degree of association;
Wherein, described first determines that device also includes:
Son determines device, for according to described distributed intelligence and other relevant informations described, determine that described pending resource is retouched
State the degree of association between information and/or its each key word comprised and every other resource description information, pending to obtain this
The accuracy information of resource description information.
15. accuracy according to claim 14 determine device, and wherein, other relevant informations described include following at least one
:
-comprise the authority of the resource described by other resource description information described of arbitrary key word in each key word described
Property;
Each key word in-described all key words is relevant to first between other resource description information comprising this key word
Degree;
Each key word in-described all key words and the second degree of association between described pending resource description information.
16. accuracy according to claim 12 determine device, and wherein, described selection device also includes:
Identifying device, be used for the network related information corresponding according to the resource described by the plurality of resource description information, it is right to come
The plurality of resource description information is identified, so that the user identifying gained is generated resource description information as described pending money
Source Description information.
17. accuracy according to claim 16 determine device, and wherein, described network related information includes following at least one
:
The link address information of the resource that-this network related information is corresponding;
The page feature information of webpage belonging to the resource that-this network related information is corresponding;
The page feature information of the webpage that resource affiliated web site corresponding to-this network related information is comprised.
18. accuracy according to claim 12 determine device, and wherein, this accuracy determines that device also includes following step
Rapid:
3rd acquisition device, it is used for obtaining multiple resource;
Clustering apparatus, for the information that self comprised according to the plurality of resource, the plurality of resource is clustered, to obtain
Obtaining one or more groups cluster resource, wherein, often group cluster resource includes one or more same or analogous resource;
Wherein, this accuracy determines that device also includes:
Construction device, for often organizing, according to described, the resource description information that same or analogous resource is corresponding, set up described money
Source Description information aggregate.
19. accuracy according to claim 12 determine device, and wherein, this accuracy determines that device also includes:
Updating device, for according to the accuracy of described pending resource description information and described resource thereof, set up or more
New resources information bank.
20. accuracy according to claim 19 determine device, and wherein, this accuracy determines that device also includes:
4th acquisition device, for obtaining the behavior relevant information relevant to user behavior;
Second determines device, for according to described behavior relevant information, determining pending resource;
Inquiry unit, for inquiring about in described resource information bank according to described pending resource, treat described in acquisition
Process the accuracy information of resource description information corresponding to resource;
Perform device, for the accuracy information according to resource description information corresponding to described pending resource, perform with described
User behavior operates accordingly.
21. accuracy according to claim 20 determine device, and wherein, described user behavior information includes following at least one
:
The type of-user operation;
The object of-user operation;
The input content inputted in-user input operation.
22. accuracy according to claim 12 determine device, described accuracy information include following at least one:
The overall accuracy of-described pending resource description information;
The accuracy of each key word that-described pending resource description information is comprised.
23. 1 kinds of computers, wherein, this computer equipment includes the accuracy as described at least one in claim 12 to 22
Determine device.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201110093719.0A CN102737059B (en) | 2011-04-14 | For determining the method for the accuracy information of resource description information, device and equipment |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201110093719.0A CN102737059B (en) | 2011-04-14 | For determining the method for the accuracy information of resource description information, device and equipment |
Publications (2)
Publication Number | Publication Date |
---|---|
CN102737059A CN102737059A (en) | 2012-10-17 |
CN102737059B true CN102737059B (en) | 2016-12-14 |
Family
ID=
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101000611A (en) * | 2006-08-29 | 2007-07-18 | 曾文均 | Method for providing and inquiry information for public by interconnection network |
CN101075942A (en) * | 2007-06-22 | 2007-11-21 | 清华大学 | Method and system for processing social network expert information based on expert value progation algorithm |
CN101089843A (en) * | 2006-06-15 | 2007-12-19 | 王刘忠 | Search method only for product or service supply information |
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101089843A (en) * | 2006-06-15 | 2007-12-19 | 王刘忠 | Search method only for product or service supply information |
CN101000611A (en) * | 2006-08-29 | 2007-07-18 | 曾文均 | Method for providing and inquiry information for public by interconnection network |
CN101075942A (en) * | 2007-06-22 | 2007-11-21 | 清华大学 | Method and system for processing social network expert information based on expert value progation algorithm |
Non-Patent Citations (1)
Title |
---|
基于关键词集合的产品信息描述与检索系统;李玉红等;《控制工程》;20050331;第12卷(第2期);第168-169页 * |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US11756245B2 (en) | Machine learning to generate and evaluate visualizations | |
DE112015002286T9 (en) | VISUAL INTERACTIVE SEARCH | |
US9633082B2 (en) | Search result ranking method and system | |
CN106415537B (en) | Locally applied search result is inserted into WEB search result | |
CN102171689B (en) | Method and system for providing search results | |
CN110020128B (en) | Search result ordering method and device | |
US9483788B2 (en) | System and method for graphically building weighted search queries | |
CN103038769B (en) | System and method for content to be directed into social network engine user | |
CN106686063A (en) | Information recommendation method and apparatus, and electronic device | |
US9088811B2 (en) | Information providing system, information providing method, information providing device, program, and information storage medium | |
CN110175895B (en) | Article recommendation method and device | |
CN106651542A (en) | Goods recommendation method and apparatus | |
CN110008397B (en) | Recommendation model training method and device | |
CN104615631B (en) | A kind of method and device of information recommendation | |
CN106407349A (en) | Product recommendation method and device | |
Yang et al. | Prototype-based image search reranking | |
CN110781377B (en) | Article recommendation method and device | |
CN112825089B (en) | Article recommendation method, device, equipment and storage medium | |
CN106682963A (en) | Recommendation system data completion method based on convex optimization local low-rank matrix approximation | |
CN109657145A (en) | Merchant searching method and device, electronic equipment and computer-readable storage medium | |
CN112232933A (en) | House source information recommendation method, device, equipment and readable storage medium | |
KR101346927B1 (en) | Search device, search method, and computer-readable memory medium for recording search program | |
CN106599291B (en) | Data grouping method and device | |
CN102760127B (en) | Method, device and the equipment of resource type are determined based on expanded text information | |
CN105512122A (en) | Ordering method and ordering device for information retrieval system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
EE01 | Entry into force of recordation of patent licensing contract |
Application publication date: 20121017 Assignee: Beijing small mutual Entertainment Technology Co., Ltd. Assignor: BEIJING BAIDU NETCOM SCIENCE AND TECHNOLOGY Co.,Ltd. Contract record no.: 2017110000013 Denomination of invention: Method, apparatus and device for determining accuracy information of resource description information Granted publication date: 20161214 License type: Exclusive License Record date: 20170705 |