CN105824938B - A kind of search method and system based on biaxial stress structure - Google Patents
A kind of search method and system based on biaxial stress structure Download PDFInfo
- Publication number
- CN105824938B CN105824938B CN201610158099.7A CN201610158099A CN105824938B CN 105824938 B CN105824938 B CN 105824938B CN 201610158099 A CN201610158099 A CN 201610158099A CN 105824938 B CN105824938 B CN 105824938B
- Authority
- CN
- China
- Prior art keywords
- event
- word
- standard
- event class
- class standard
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/951—Indexing; Web crawling techniques
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The present invention relates to a kind of search method and system based on biaxial stress structure, the present invention is mapped by level-one mapping and second level, on the one hand solve the problems, such as that user cannot position oneself actual retrieval demand since legal knowledge is deficient very well, on the other hand solve the problems, such as due to various regions law court administrative case case by determination and statement it is not the same and cause lawyer's professional services evaluation it is incomplete, the present invention is on the basis of above-mentioned two aspects mapping, it further realizes user search demand and is docked with object is retrieved, to reach precise positioning user search demand, improve the accuracy of search result, and improve data search efficiency.
Description
Technical field
The invention belongs to data retrieval technology field, in particular to a kind of search method and system based on biaxial stress structure.
Background technique
Legal profession threshold is high, strongly professional, and the legal knowledge of ordinary user is relatively deficient, this is just directly limited
Means of communication and communicative channel between party and law supplier.Simultaneously as adat service industry information source
In a jumble, party generally lacks the effective way for finding suitable lawyer, and understanding is looked for by the lawyer's office for usually requiring entity
People or the lawyer that is recommended of rule, both taken a substantial amount of time and energy, the uncertain answer that can obtain oneself satisfaction.
For internet legal services platform now in view of the above-mentioned problems, by lawyer's resource consolidation to network, user can
The lawyer's progress legal advice or case commission for meeting oneself demand are searched directly on platform.However, due to party user
It is generally all the absence of knowledge of specialty law, and legal industry is professional very strong, even if which results in having internet
In the case that legal services platform provides lawyer's resource online, user also due to the scarcity of legal knowledge and cannot position well
Suitable lawyer is found in the actual demand of oneself.In addition, various regions law court administrative case case by determination and statement it is not the same,
Had a certain difference with most Supreme Court's case by dividing, thus can not comprehensively and study plot according to the document datas such as judgement document to rule
Shi Jinhang is objectively evaluated, and leading to lawyer's search result, there is also certain errors.
In order to solve the above-mentioned technical problem, therefore technical solution of the present invention is proposed.
Summary of the invention
The object of the present invention is to provide a kind of search method and system based on biaxial stress structure, the present invention are mapped by level-one
It is mapped with second level, asking for oneself actual retrieval demand cannot be positioned very well by the one hand solving user since legal knowledge is deficient
Topic, on the other hand solve due to various regions law court administrative case case by determination and statement it is not the same and lead to lawyer's operation feelings
Condition evaluates incomplete problem, the present invention on the basis of establishing above-mentioned two aspects mapping, further realize user search demand with
The docking of object is retrieved, to reach precise positioning user search demand, so that improving search result accuracy, and improves data inspection
Rope efficiency.
To solve the above problems, one aspect of the present invention provides a kind of search method based on biaxial stress structure, it is described
Method includes: to carry out keyword extraction to the event information that client obtains, and is obtained comprising at least one event non-standard feature
The set of word；Based on the level-one mapping relations between event non-standard Feature Words and event class standard statement word, to the event
Each of the non-standard feature set of words event non-standard Feature Words are standardized, to obtain each non-rule of the event
At least one corresponding event class standard of model Feature Words states word, forms event class standard and states set of words；Traverse the event
Class standard states event class standard all in set of words and states word, states the rank between word according to event class standard and is subordinate to pass
It is that membership analysis two-by-two is successively unfolded to the event class standard statement word in the set, to obtain event class standard statement
The subset of set of words；Based on the second level mapping relations between document characteristic key words and event class standard statement word, to the thing
Each of the part class standard statement lexon concentration event class standard statement word successively retrieves second level mapping relations, based on retrieval knot
Fruit generates the displayed page comprising at least one document characteristic key words.
Wherein, the client obtains the step of event information comprising: it is obtained recently from search engine server
Retrieval log in the setting time length and retrieval log that will acquire is as event information；Or by this user in client
The event information of input is as event information.
Wherein, the level-one mapping relations based between event non-standard Feature Words and event class standard statement word are right
Each of the event non-standard feature set of words event non-standard Feature Words are standardized, to obtain each event
The step of at least one corresponding event class standard of non-standard Feature Words states word, forms event class standard statement set of words, packet
It includes: establishing the level-one mapping relations between event non-standard Feature Words and event class standard statement word, be saved in standardization statement
In library；Traversal event non-standard feature set of words { FnIn all event non-standard Feature Words, it is non-to obtain each event
Specification features word FnAt least one corresponding event class standard states word Sm, form the thing comprising multiple event class standards statement word
Part class standard states set of words { Sm},n≥m≥1。
Wherein, the method also includes: if from standardization statement library in do not obtain some described event non-standard Feature Words
FiCorresponding event class standard states word Si, then the standardization can be filtered by Text similarity computing method and states library, obtained
To with event non-standard Feature Words FiApproximate event class standard states word Sj, 1≤i≤n, 1≤j≤n；It is non-to establish the event
Specification features word FiWord S is stated with the approximate event class standardjBetween level-one mapping relations, by the event non-standard feature
Word and the approximate event class standard state word SjBetween level-one mapping relations be saved in standardization statement library in.
Wherein, the step for establishing the level-one mapping relations between event non-standard Feature Words and event class standard statement word
Suddenly comprising: the user behavior data that client is got is captured, therefrom extracts and obtains at least one event non-standard spy
Levy word；Based at least one described event non-standard Feature Words, establish at least one described event non-standard Feature Words with it is described
Event class standard states one-to-one, the one-to-many or many-to-one level-one mapping relations between word；Wherein, the event class standard
Stating word is the corresponding case type of the event non-standard Feature Words.
Wherein, described that rank membership between word is stated to event class standard statement word according to event class standard
Event class standard statement word in set carries out membership analysis, to obtain the subset of event class standard statement set of words
Step comprising: when there are an event class standards to state word S in event class standard statement set of wordspOr it is multiple identical
Event class standard state word SpWhen, it is determined that the event class standard states word SpAs event class standard statement set of words
Subset.
Wherein, described that rank membership between word is stated to event class standard statement word according to event class standard
Event class standard statement word in set carries out membership analysis, to obtain the subset of event class standard statement set of words
Step comprising: it is when stating word there are multiple and different event class standards in event class standard statement set of words, then right
Membership analysis two-by-two is successively unfolded in event class standard statement word in the set；If event class standard states word SjWith event
Class standard states word Sj+1Belong to coordination at the same level, it is determined that { Sj,Sj+1It is the subset that the event class standard states set of words；
If event class standard states word SjWord S is stated with event class standardj+1Belong to covering relation at the same level, and SjCover Sj+1, it is determined that
{SjIt is the subset that the event class standard states set of words；If event class standard states word SjWord is stated comprising event class standard
Sj+1, then { Sj+1It is the subset that the event class standard states set of words；If event class standard states word SjIt is contained in event category
Quasi- statement word Sj+1, then { SjIt is the subset that the event class standard states set of words.
Wherein, the second level mapping relations based between document characteristic key words and event class standard statement word, to institute
It states each of the event class standard statement lexon concentration event class standard statement word and successively retrieves second level mapping relations, based on inspection
Hitch fruit generates the displayed page step comprising at least one document characteristic key words comprising: establish document feature pass
Second level mapping relations between keyword and event class standard statement word are saved in standardization statement library；Based on document feature critical
Second level mapping relations between word and event class standard statement word traverse all of the event class standard statement lexon concentration
The event class standard states word, obtains the corresponding document characteristic key words of each event class standard statement word；To described
Document characteristic key words carry out statistical disposition, generate the displayed page comprising at least one document characteristic key words.
Wherein, the step of second level mapping relations established between document characteristic key words and event class standard statement word
It include: to carry out keyword extraction to all documents in document database to obtain document characteristic key words；It is special to establish the document
Levy one-to-one or many-to-one second level mapping relations between keyword and event class standard statement word.
Another aspect of the present invention provides a kind of searching system based on biaxial stress structure, the system comprises: it is crucial
Word extraction module, the event information for obtaining to client carry out keyword extraction, obtain comprising the non-rule of at least one event
The set of model Feature Words；Standard scale predicate obtains module, for stating word based on event non-standard Feature Words and event class standard
Between level-one mapping relations, each thing in the event non-standard feature set of words is obtained to the keyword-extraction module
Part non-standard Feature Words are standardized, to obtain at least one corresponding event category of each event non-standard Feature Words
Quasi- statement word, forms event class standard and states set of words；Membership analysis module, for traversing the event class standard statement
All event class standards state word in set of words, state the rank membership between word to the set according to event class standard
In event class standard statement word be successively unfolded two-by-two membership analysis, with obtain the event class standard statement set of words son
Collection；Search result generation module is closed for being mapped based on the second level between document characteristic key words and event class standard statement word
System successively retrieves second level mapping to each of the event category fiducial mark statement lexon concentration event class standard statement word and closes
System generates the displayed page comprising at least one document characteristic key words based on search result.
Wherein, the system also includes: event informations to obtain module, nearest for obtaining from search engine server
Retrieval log in the setting time length and retrieval log that will acquire is as event information；Or by this user in client
The event information of input is as event information.
Wherein, it includes: that unit is established in level-one mapping that the standard scale predicate, which obtains module, special for establishing event non-standard
The level-one mapping relations between word and event class standard statement word are levied, are saved in standardization statement library；Non-standard Feature Words time
Unit is gone through, for traversing event non-standard feature set of words { Fn, obtain each event non-standard Feature Words FnIt is corresponding extremely
A few event class standard states word Sm, form the event class standard comprising multiple event class standards statement word and state set of words
{Sm, n >=m >=1.
Wherein, the Feature Words Traversal Unit further comprises: approximation statement word acquiring unit, for working as standard scale predicate
Traversal Unit does not obtain some described event non-standard Feature Words F from standardization statement libraryiCorresponding event class standard statement
Word Si, the standardization is filtered by Text similarity computing method and states library, is obtained and event non-standard Feature Words FiIt is approximate
Event class standard state word Sj, 1≤i≤n, 1≤j≤n；Level-one map updating unit, it is special for establishing the event non-standard
Levy word FiWord S is stated with the approximate event class standardjBetween level-one mapping relations, and by the event non-standard Feature Words with
The approximate event class standard states word SjBetween level-one mapping relations be saved in standardization statement library in.
Wherein, level-one mapping establishes unit and further comprises: user behavior analysis unit is obtained for capturing client
The user behavior data got therefrom extracts and obtains at least one described event non-standard Feature Words；The level-one mapping is established
Unit, for based at least one described event non-standard Feature Words, establish at least one described event non-standard Feature Words with
One-to-one, one-to-many or many-to-one level-one mapping relations between the event class standard statement word；Wherein, the event class
Standard scale predicate is the corresponding case type of the event non-standard Feature Words.
Wherein, when the membership analysis module detects that there are a things in the event class standard statement set of words
Part class standard states word SpOr multiple identical event class standards state word SpWhen, it is determined that the event class standard states word SpMake
The subset of set of words is stated for the event class standard.
Wherein, when the membership analysis module detect in event class standard statement set of words there are it is multiple not
When same event class standard states word, then membership point two-by-two is successively unfolded to the event class standard statement word in the set
Analysis, the membership analysis module specifically execute following operation: if event class standard states word SjIt is stated with event class standard
Word Sj+1Belong to coordination at the same level, it is determined that { Sj,Sj+1It is the subset that the event class standard states set of words；If event category
Quasi- statement word SjWord S is stated with event class standardj+1Belong to covering relation at the same level, and SjCover Sj+1, it is determined that { SjIt is the event
The subset of class standard statement set of words；If event class standard states word SjWord S is stated comprising event class standardj+1, then { Sj+1Be
The subset of event class standard statement set of words；If event class standard states word SjIt is contained in event class standard statement word Sj+1, then
{SjIt is the subset that the event class standard states set of words.
Wherein, the search result generation module includes: that unit is established in second level mapping, for establishing document characteristic key words
With the second level mapping relations between event class standard statement word, it is saved in standardization statement library；Standard scale predicate Traversal Unit is used
In based on the second level mapping relations between document characteristic key words and event class standard statement word, the event class standard table is traversed
All event class standards in predicate subset state word, obtain the corresponding document of each event class standard statement word
Characteristic key words；Displayed page generation unit is generated for carrying out statistical disposition to the document characteristic key words comprising described
The displayed page of at least one document characteristic key words.
Wherein, it includes: document keyword extracting unit that unit is established in the second level mapping, for in document database
All documents carry out keyword extraction and obtain document characteristic key words；Unit is established in the second level mapping, for establishing the text
One-to-one or many-to-one second level mapping relations between book characteristic key words and event class standard statement word.
On the one hand a kind of search method and system based on biaxial stress structure provided by the present invention solves user due to method
The problem of restraining knowledge scarcity and cannot positioning oneself actual retrieval demand very well, on the other hand solves due to the administration of various regions law court
Case case by determination and statement it is not the same and lawyer's professional services is caused to evaluate incomplete problem.By establishing user terminal
Level-one between non-standard Feature Words and standard scale predicate maps and based between document key feature word and standard scale predicate
Second level mapping further realizes user search demand and docks with retrieval object, to reach precise positioning user search demand, makes
Search result accuracy must be improved, and improves data search efficiency.
Detailed description of the invention
Fig. 1 is the flow diagram of the search method of the invention based on biaxial stress structure；
Fig. 2 is the flow diagram of step S2 of the invention；
Fig. 3 is the flow diagram of step S21 of the invention；
Fig. 4 is the flow diagram of step S3 of the invention；
Fig. 5 is the flow diagram of step S4 of the invention；
Fig. 6 is the flow diagram of step S41 of the invention；
Fig. 7 is the structural schematic diagram of the searching system of the invention based on biaxial stress structure；
Fig. 8 is the structural schematic diagram that standard scale predicate of the invention obtains module；
Fig. 9 is the structural schematic diagram of Feature Words Traversal Unit of the invention；
Figure 10 is the structural schematic diagram of search result generation module of the invention.
Specific embodiment
In order to make the objectives, technical solutions and advantages of the present invention clearer, With reference to embodiment and join
According to attached drawing, the present invention is described in more detail.It should be understood that these descriptions are merely illustrative, and it is not intended to limit this hair
Bright range.In addition, in the following description, descriptions of well-known structures and technologies are omitted, to avoid this is unnecessarily obscured
The concept of invention.
Fig. 1 is the flow diagram of the search method of the invention based on biaxial stress structure.
As shown in Figure 1, the search method of the invention based on biaxial stress structure comprising:
Step S1 carries out keyword extraction to the event information that client obtains, obtains comprising the non-rule of at least one event
The set of model Feature Words.
In this step, for user in client incoming event information, the event information is usually that user retrieves needs
The event form of presentation for the non-standard that event carries out, client carry out keyword extraction to the event information, obtain comprising one
The set of a or multiple event non-standard Feature Words.
For example, therefrom can only extract the non-rule of event when the event information of user's input is " I will divorce "
Model Feature Words are " divorce ", then non-standard feature set of words is { " divorce " }.For another example when the event information of user's input is
" polygamous divorce issue ", the keyword that therefrom can extract are " polygamy " and " divorce ", then non-standard Feature Words
Collection is combined into { " polygamy ", " divorce " }.
Specifically, the client obtains the step of event information, further comprising: obtaining from search engine server
It takes the retrieval log in nearest setting time length and the retrieval log that will acquire is as event information；Or by this user
In the event information that client inputs as event information.For example, the retrieval log in nearest 2 days is obtained, by the retrieval log
As event information；Or this event information for inputting user in client is as current event information.
Step S2, based on the level-one mapping relations between event non-standard Feature Words and event class standard statement word, to institute
It states each of the event non-standard feature set of words event non-standard Feature Words to be standardized, to obtain each thing
At least one corresponding event class standard of part non-standard Feature Words states word, forms event class standard and states set of words.
In this step, the level-one mapping relations between word are stated based on event non-standard Feature Words and event class standard, into
One step is standardized each of the event non-standard feature set of words event non-standard Feature Words, standardized
Purpose is to obtain corresponding at least one the event class standard statement word of each event non-standard Feature Words, forms event category
Quasi- statement set of words.
For example, accepting above-mentioned example, reflected based on the level-one between event non-standard Feature Words and event class standard statement word
Relationship is penetrated, it may corresponding event class standard statement word " divorce dispute ", " divorce to obtain event non-standard Feature Words " divorce "
Liability for damage dispute afterwards ", " property dispute after divorce ", then can establish event non-standard Feature Words and event class standard statement word it
Between one-to-many mapping relations.
For another example above-mentioned example is accepted, based on the level-one between event non-standard Feature Words and event class standard statement word
Mapping relations, the possible corresponding event class standard statement word of event non-standard Feature Words " polygamy " is that " annulment of marriage entangles
Confusingly "；The possible corresponding event class standard statement word of event non-standard Feature Words " divorce " includes " divorce dispute ", " damages after divorce
Evil responsibility dispute ", " property dispute after divorce " etc. can then establish between event non-standard Feature Words and event class standard statement word
One-to-one or one-to-many mapping relations.
The specific step that executes see below example shown in Fig. 2.
Step S3 traverses event class standard statement word all in the event class standard statement set of words, according to event
Rank membership between class standard statement word is successively unfolded to be subordinate to pass two-by-two to the event class standard statement word in the set
System's analysis, to obtain the subset of event class standard statement set of words.
In this step, event class standard statement word all in the event class standard statement set of words is traversed, according to thing
Rank membership between part class standard statement word is successively unfolded to be subordinate to two-by-two to the event class standard statement word in the set
Relationship analysis, to obtain the subset of event class standard statement set of words.
In the present invention, the rank membership between the event class standard statement word includes that event class standard statement word is deposited
Coordination at the same level, covering relation at the same level and comprising/be contained in relationship.Specifically, the grade between event class standard statement word
Other membership includes multiple event class standard statement word major class, and each event class standard statement word major class includes multiple event classes
Standard scale predicate group.Event class standard states between word major class, there is covering at the same level between event class standard statement word group
With coordination at the same level, event class standard statement word major class and event class standard statement word group between exist comprising/be contained in
Relationship.
For example, it is assumed that event class standard statement set of words includes { S1, S2, S3, if event class standard states word S1With event
Class standard states word S2Belong to coordination at the same level, then { S1,S2It is the subset that the event class standard states set of words；At this point, if
Event class states word S3Word S is stated with event class2Coordination at the same level is also belonged to, then subset is updated to { S1, S2, S3}；Continue into
It acts part class and states word S3Word S is stated with event class1The analysis of relationship, if event class states word S3Word S is stated with event class1Also
Belong to coordination at the same level, then subset is determined as { S1, S2, S3}；If event class states word S3Word S is stated with event class1Belong to same
Event class statement word S ought occur simultaneously in grade covering relation, i.e. legal provisions3Word S is stated with event class1When, case class is regarded as
S3, then subset is determined as { S2, S3}。
Example in undertaking, when event class standard states word S1Word S is stated with event class standard2When belonging to coordination at the same level, if
Event class states word S3Word S is stated comprising event class2, then using short-circuit logic, subset is determined as { S1, S2}；If event class is stated
Word S3It is contained in event class statement word S2, then using short-circuit logic, subset is determined as { S3}。
For example, accept upper example, event class standard statement word set is combined into that { " A divorce dispute ", " liability for damage entangles after B divorce
Confusingly ", " property dispute after C divorce ", " dispute of D annulment of marriage " }, successively expansion membership analysis two-by-two, legal provisions standard
It states word A peer's coverage criteria and states word B and standard scale predicate C, then subset is { " A divorce dispute ", " D wedding after 2 analyses
The invalid dispute of relation by marriage " }；Continue analytical standard and states word A and standard scale predicate D, legal provisions standard scale predicate D peer coverage criteria
Word A is stated, then set is updated to { " dispute of D annulment of marriage " }.
The specific step that executes see below example shown in Fig. 4.
Step S4, based on the second level mapping relations between document characteristic key words and event class standard statement word, to described
Each of the event class standard statement lexon concentration event class standard statement word successively retrieves second level mapping relations, based on retrieval
As a result the displayed page comprising at least one document characteristic key words is generated.
The specific step that executes see below example shown in Fig. 5.
Fig. 2 is the flow diagram of step S2 of the invention.
As shown in Figure 2, wherein abovementioned steps S2 further comprises:
Step S21 establishes the level-one mapping relations between event non-standard Feature Words and event class standard statement word, saves
Into standardization statement library.
The specific step that executes see below example shown in Fig. 3.
Step S22 traverses event non-standard feature set of words { FnIn all non-standard Feature Words, obtain each described
Event non-standard Feature Words FnAt least one corresponding event class standard states word Sm, formed and stated comprising multiple event class standards
The event class standard of word states set of words { Sm},n≥m≥1。
In this step, there are three kinds of situations: the first situation: the corresponding event category of an event non-standard Feature Words
Quasi- statement word；Second situation: an event non-standard Feature Words correspond to multiple and different event class standard statement words；It is multiple not
The corresponding event class standard of same non-standard Feature Words states word.
Abovementioned steps S22 is accepted, the method also includes:
Step S23, if not obtaining some described event non-standard Feature Words F from standardization statement libraryiCorresponding event
Class standard states word Si, then the standardization can be filtered by Text similarity computing method and states library, obtained and the event
Non-standard Feature Words FiApproximate event class standard states word Sj, 1≤i≤n, 1≤j≤n；
It should be understood that Text similarity computing method include but is not limited to text similarity co sinus vector included angle,
The methods of Simhash algorithm can determine text by the methods of the co sinus vector included angle of text similarity, Simhash algorithm
Whether height is approximate for this content, and details are not described herein again.
Step S24 establishes event non-standard Feature Words FiWord S is stated with the approximate event class standardjBetween level-one
The event non-standard Feature Words and the approximate event class standard are stated word S by mapping relationsjBetween level-one mapping relations protect
It is stored in standardization statement library.
Fig. 3 is the flow diagram of step S21 of the invention.
Wherein, abovementioned steps S21 further comprises:
Step S211, captures the user behavior data that client is got, and therefrom extracts and obtains at least one described event
Non-standard Feature Words.
Step S212 establishes at least one described event non-standard based at least one described event non-standard Feature Words
One-to-one, one-to-many or many-to-one level-one mapping relations between Feature Words and event class standard statement word；Wherein, institute
Stating event class standard statement word is the corresponding case type of the event non-standard Feature Words.
For example, therefrom extracting when user behavior data " divorce " that client captures, " bigamy " and obtaining two non-rule
Model Feature Words " divorce ", " bigamy ", " consanguineous marriage ".Wherein, " divorce " possible corresponding event class standard statement word is " divorce
Dispute ", " liability for damage dispute after divorce ", " property dispute after divorce "；" bigamy " may corresponding event class standard statement word
For " annulment of marriage dispute "；" consanguineous marriage " possible corresponding event class standard statement word is " annulment of marriage dispute "；It can then build
One-to-one, one-to-many or many-to-one mapping relations between vertical event non-standard Feature Words and event class standard statement word.
Fig. 4 is the flow diagram of step S3 of the invention.
As shown in figure 4, above mentioned step S3 further comprise:
Step S31, when there are an event class standards to state word S in event class standard statement set of wordspOr it is multiple
Identical event class standard states word SpWhen, it is determined that the event class standard states word SpWord set is stated as the event class standard
The subset of conjunction.
For example, when event class standard statement word set is combined into { S1When, it is determined that the event class standard states word S1For the event
The subset of class standard statement set of words.For another example when event class standard statement word set is combined into { S1, S1, and S1, S1It is identical
When event class standard states word, it is determined that { S1It is the subset that the event class standard states set of words.
Step S32, when stating word there are multiple and different event class standards in event class standard statement set of words,
Membership analysis two-by-two is successively then unfolded to the event class standard statement word in the set.
Specifically, this step includes three kinds of situations: coordination at the same level, covering relation at the same level, comprising/be contained in relationship:
Step S321, if event class standard states word SjWord S is stated with event class standardj+1Belong to coordination at the same level, then
Determine { Sj,Sj+1It is the subset that the event class standard states set of words.
Step S322, if event class standard states word SjWord S is stated with event class standardj+1Belong to covering relation at the same level, i.e.,
SjCover Sj+1, it is determined that { SjIt is the subset that the event class standard states set of words.
Step S323, if event class standard states word SjWord S is stated comprising event class standardj+1, then { Sj+1It is the event
The subset of class standard statement set of words；
If event class standard states word SjIt is contained in event class standard statement word Sj+1, then { SjIt is the set event class standard
State the subset of set of words.
Fig. 5 is the flow diagram of step S4 of the invention.
As shown in figure 5, abovementioned steps S4 includes:
Step S41 establishes the second level mapping relations between document characteristic key words and event class standard statement word, is saved in
Standardization statement library.
It should be understood that document characteristic key words refer to the case type occurred in document in the present invention.
Step S42 traverses institute based on the second level mapping relations between document characteristic key words and event class standard statement word
All event class standard statement words that event class standard statement lexon is concentrated are stated, each event class standard table is obtained
The corresponding document characteristic key words of predicate.
Step S43 carries out statistical disposition to the document characteristic key words, generates comprising at least one described document feature
The displayed page of keyword.
Fig. 6 is the flow diagram of step S41 of the invention.
As shown in fig. 6, abovementioned steps S41 includes:
Step S411 carries out keyword extraction to all documents in document database and obtains document characteristic key words.
It should be understood that document describes lawyer's processed all case information during operation in the present invention, with
And the personal information of lawyer.
Step S412 establishes one-to-one or many-one between the document characteristic key words and event class standard statement word
Second level mapping relations；Wherein, the document characteristic key words are case type, law court and/or lawyer.
The step includes two kinds of situations: the first situation: the corresponding event class standard table of a document characteristic key words
Predicate；Second situation: the corresponding event class standard of multiple document characteristic key words (i.e. multiple documents) states word.
As described above, detailing a kind of search method based on biaxial stress structure of the invention, the present invention is mapped by level-one
It is mapped with second level, asking for oneself actual retrieval demand cannot be positioned very well by the one hand solving user since legal knowledge is deficient
Topic, on the other hand solve due to various regions law court administrative case case by determination and statement it is not the same and lead to lawyer's operation feelings
Condition evaluates incomplete problem, and the present invention can further realize user search demand on the basis of above-mentioned two aspects mapping
So that improving search result accuracy, and data are improved with docking for retrieval object to reach precise positioning user search demand
Recall precision.
Fig. 7 is the structural schematic diagram of the searching system of the invention based on biaxial stress structure.
As shown in fig. 7, the searching system based on biaxial stress structure of the invention, the system comprises: keyword extraction
Module 10, standard scale predicate obtain module 20, membership analysis module 30, search result generation module 40 and event information and obtain
Modulus block 50.
Keyword-extraction module 10, the event information for obtaining to client carry out keyword extraction, obtain comprising extremely
The set of few event non-standard Feature Words.
It obtains module 50 the system also includes: event information to connect with the keyword-extraction module 10, for from searching
The retrieval log in nearest setting time length is obtained on rope engine server and the retrieval log that will acquire is believed as event
Breath；Or the event information for inputting this user in client is as event information.
It specifically describes referring to abovementioned steps S1.
Standard scale predicate obtains module 20 and connect with the keyword-extraction module 10, for being based on event non-standard feature
Level-one mapping relations between word and event class standard statement word, obtain the non-rule of the event to the keyword-extraction module 10
Each event non-standard Feature Words in model feature set of words are standardized, to obtain each event non-standard Feature Words
At least one corresponding event class standard states word, forms event class standard and states set of words.
It specifically describes referring to abovementioned steps S2.
Membership analysis module 30 obtains module 20 with the standard scale predicate and connect, for traversing the event category
All event class standards state word in quasi- statement set of words, state the rank membership pair between word according to event class standard
Membership analysis two-by-two is successively unfolded in event class standard statement word in the set, to obtain event class standard statement word set
The subset of conjunction.
Specifically, when the membership analysis module 30 detects that there are one in the event class standard statement set of words
A event class standard states word SpOr multiple identical event class standards state word SpWhen, it is determined that the event class standard states word
SpSubset as event class standard statement set of words.
When the membership analysis module 30 detects that there are multiple and different in the event class standard statement set of words
Event class standard state word when, then in the set event class standard statement word be successively unfolded two-by-two membership analysis,
The membership analysis module 30 specifically executes following operation:
If event class standard states word SjWord S is stated with event class standardj+1Belong to coordination at the same level, it is determined that { Sj,
Sj+1It is the subset that the event class standard states set of words.
If event class standard states word SjWord S is stated with event class standardj+1Belong to covering relation at the same level, i.e. SjCover Sj+1,
Then determine { SjIt is the subset that the event class standard states set of words.
If event class standard states word SjWord S is stated comprising event class standardj+1, then { Sj+1It is that the event class standard is stated
The subset of set of words.
If event class standard states word SjIt is contained in event class standard statement word Sj+1, then { SjIt is that the event class standard is stated
The subset of set of words.
It specifically describes referring to above mentioned step S3, step S31, step S32.
Search result generation module 40 is connect with the membership analysis module 30, for being based on document characteristic key words
It is described to each of event category fiducial mark statement lexon concentration with the second level mapping relations between event class standard statement word
Event class standard statement word successively retrieves second level mapping relations, is generated based on search result comprising at least one described document feature
The displayed page of keyword.
Fig. 8 is the structural schematic diagram that standard scale predicate of the invention obtains module.
As shown in Figure 8, wherein the standard scale predicate obtains module 20 and further comprises: unit 201 is established in level-one mapping
With non-standard Feature Words Traversal Unit 202.
Unit 201 is established in level-one mapping, for establishing between event non-standard Feature Words and event class standard statement word
Level-one mapping relations are saved in standardization statement library.
It specifically describes referring to abovementioned steps S21.
Non-standard Feature Words Traversal Unit 202 is established unit 201 with level-one mapping and is connect, for traversing the non-rule of event
Model feature set of words { Fn, obtain each event non-standard Feature Words FnAt least one corresponding event class standard states word
Sm, form the event class standard comprising multiple event class standards statement word and state set of words { Sm, n >=m >=1.
It specifically describes referring to abovementioned steps S22.
Fig. 9 is the structural schematic diagram of Feature Words Traversal Unit of the invention.
As shown in figure 9, the Feature Words Traversal Unit 202 further comprises: approximation statement word acquiring unit 2020 and one
Grade map updating unit 2021.
Approximate statement word acquiring unit 2020, for being stated in library not when standard scale predicate Traversal Unit 202 from standardization
Obtain some described event non-standard Feature Words FiCorresponding event class standard states word Si, then pass through Text similarity computing side
Method filters the standardization and states library, obtains and event non-standard Feature Words FiApproximate event class standard states word Sj, 1≤i
≤ n, 1≤j≤n.
It specifically describes referring to abovementioned steps S23.
Level-one map updating unit 2021 is connected with the approximate statement word acquiring unit 2020, non-for establishing the event
Specification features word FiWord S is stated with the approximate event class standardjBetween level-one mapping relations, and it is event non-standard is special
It levies word and the approximate event class standard states word SjBetween level-one mapping relations be saved in standardization statement library in.
It specifically describes referring to abovementioned steps S24.
As shown in figure 8, unit 201 is established in level-one mapping further comprises: user behavior analysis unit 2010.
User behavior analysis unit 2010 is established unit 201 with level-one mapping and is connect, and obtains for capturing client
The user behavior data arrived therefrom extracts and obtains at least one described event non-standard Feature Words.
It specifically describes referring to abovementioned steps S211.
Unit 201 is established in level-one mapping, for based at least one described event non-standard Feature Words, described in foundation
One-to-one, one-to-many or many-to-one one between at least one event non-standard Feature Words and event class standard statement word
Grade mapping relations；Wherein, the event class standard statement word is the corresponding case type of the event non-standard Feature Words.
It specifically describes referring to abovementioned steps S212.
Figure 10 is the structural schematic diagram of search result generation module of the invention.
As shown in Figure 10, the search result generation module 40 includes: that unit 401, standard scale predicate are established in second level mapping
Traversal Unit 402 and displayed page generation unit 403.
Unit 401 is established in second level mapping, two for establishing between document characteristic key words and event class standard statement word
Grade mapping relations are saved in standardization statement library.
It specifically describes referring to abovementioned steps S41.
Standard scale predicate Traversal Unit 402 is established unit 402 with second level mapping and is connect, for being closed based on document feature
Second level mapping relations between keyword and event class standard statement word traverse all of the event class standard statement lexon concentration
The event class standard state word, obtain the corresponding document characteristic key words of each event class standard statement word.
It specifically describes referring to abovementioned steps S42.
Displayed page generation unit 403 is connect with the standard scale predicate Traversal Unit 402, for the document feature
Keyword carries out statistical disposition, generates the displayed page comprising at least one document characteristic key words.
It specifically describes referring to abovementioned steps S43.
As shown in Figure 10, it includes: document keyword extracting unit 4010 that unit 401 is established in the second level mapping.
Document keyword extracting unit 4010 is established unit 401 with second level mapping and is connect, for document database
In all documents carry out keyword extraction obtain document characteristic key words.
It specifically describes referring to abovementioned steps S411.
Unit 401 is established in the second level mapping, states word for establishing the document characteristic key words and event class standard
Between one-to-one or many-to-one second level mapping relations；Wherein, the document characteristic key words be case type, law court and/
Or lawyer.
It specifically describes referring to abovementioned steps S412.
As described above, detailing a kind of searching system based on biaxial stress structure of the invention, the present invention is mapped by level-one
It is mapped with second level, asking for oneself actual retrieval demand cannot be positioned very well by the one hand solving user since legal knowledge is deficient
Topic, on the other hand solve due to various regions law court administrative case case by determination and statement it is not the same and lead to lawyer's operation feelings
Condition evaluates incomplete problem, and the present invention can further realize user search demand on the basis of above-mentioned two aspects mapping
So that improving search result accuracy, and data are improved with docking for retrieval object to reach precise positioning user search demand
Recall precision.
It should be understood that above-mentioned specific embodiment of the invention is used only for exemplary illustration or explains of the invention
Principle, but not to limit the present invention.Therefore, that is done without departing from the spirit and scope of the present invention is any
Modification, equivalent replacement, improvement etc., should all be included in the protection scope of the present invention.In addition, appended claims purport of the present invention
Covering the whole variations fallen into attached claim scope and boundary or this range and the equivalent form on boundary and is repairing
Change example.
Claims (18)
1. a kind of search method based on biaxial stress structure, which is characterized in that the described method includes:
Keyword extraction is carried out to the event information that client obtains, obtains the collection comprising at least one event non-standard Feature Words
It closes；
Based on the level-one mapping relations between event non-standard Feature Words and event class standard statement word, to the event non-standard
Each of the feature set of words event non-standard Feature Words are standardized, to obtain each event non-standard feature
At least one corresponding event class standard of word states word, forms event class standard and states set of words；
Event class standard statement word all in the event class standard statement set of words is traversed, word is stated according to event class standard
Between rank membership in the set event class standard statement word be successively unfolded two-by-two membership analysis, to obtain
The subset of event class standard statement set of words；
Based on the second level mapping relations between document characteristic key words and event class standard statement word, to the event class standard table
Each of the predicate subset event class standard statement word successively retrieves second level mapping relations, includes based on search result generation
The displayed page of at least one document characteristic key words.
2. according to the method described in claim 1, wherein, the step of client obtains event information comprising:
The retrieval log in nearest setting time length is obtained from search engine server and the retrieval log that will acquire is made
For event information；Or the event information for inputting this user in client is as event information.
3. -2 described in any item methods according to claim 1, wherein described to be based on event non-standard Feature Words and event category
Level-one mapping relations between quasi- statement word, to each of the event non-standard feature set of words event non-standard Feature Words
It is standardized, to obtain corresponding at least one the event class standard statement word of each event non-standard Feature Words, is formed
Event class standard states the step of set of words, comprising:
The level-one mapping relations between event non-standard Feature Words and event class standard statement word are established, standardization statement is saved in
In library；
Traversal event non-standard feature set of words { FnIn all non-standard Feature Words, it is special to obtain each event non-standard
Levy word FnAt least one corresponding event class standard states word Sm, form the event category comprising multiple event class standards statement word
Quasi- statement set of words { Sm},n≥m≥1。
4. according to the method described in claim 3, wherein, the method also includes:
If not obtaining some described event non-standard Feature Words F from standardization statement libraryiCorresponding event class standard states word
Si, then the standardization is filtered by Text similarity computing method and states library, obtained and event non-standard Feature Words FiIt is approximate
Event class standard state word Sj, 1≤i≤n, 1≤j≤n；
Establish event non-standard Feature Words FiWord S is stated with the approximate event class standardjBetween level-one mapping relations, will
The event non-standard Feature Words and the approximate event class standard state word SjBetween level-one mapping relations be saved in standardization
It states in library.
It is described to establish event non-standard Feature Words and event class standard states word 5. according to the method described in claim 4, wherein
Between level-one mapping relations the step of comprising:
The user behavior data that client is got is captured, therefrom extracts and obtains at least one described event non-standard Feature Words；
Based at least one described event non-standard Feature Words, at least one described event non-standard Feature Words and the thing are established
Part class standard states one-to-one, the one-to-many or many-to-one level-one mapping relations between word；
Wherein, the event class standard statement word is the corresponding case type of the event non-standard Feature Words.
6. method according to claim 1,2 or 4, wherein the rank person in servitude stated according to event class standard between word
Membership analysis two-by-two is successively unfolded to the event class standard statement word in the set in category relationship, to obtain the event class standard
The step of stating the subset of set of words comprising:
When there are an event class standards to state word S in event class standard statement set of wordspOr multiple identical event categories
Quasi- statement word SpWhen, it is determined that the event class standard states word SpSubset as the set.
7. method according to claim 1,2 or 4, wherein the rank person in servitude stated according to event class standard between word
Category relationship carries out membership analysis to the event class standard statement word in event class standard statement set of words, to be somebody's turn to do
Event class standard states the step of subset of set of words comprising:
When stating word there are multiple and different event class standards in event class standard statement set of words, then in the set
Event class standard statement word be successively unfolded two-by-two membership analysis；
If event class standard states word SjWord S is stated with event class standardj+1Belong to coordination at the same level, it is determined that { Sj,Sj+1Be
The subset of event class standard statement set of words；
If event class standard states word SjWord S is stated with event class standardj+1Belong to covering relation at the same level, and event class standard is stated
Word SjCovering event class standard states word Sj+1, it is determined that { SjIt is the subset that the event class standard states set of words；
If event class standard states word SjWord S is stated comprising event class standardj+1, then { Sj+1It is that the event class standard states word set
The subset of conjunction；
If event class standard states word SjIt is contained in event class standard statement word Sj+1, then { SjIt is that the event class standard states word set
The subset of conjunction.
8. method according to claim 1,2 or 4, wherein described to be based on document characteristic key words and event class standard table
Second level mapping relations between predicate state word to each of the event class standard statement lexon concentration event class standard
Second level mapping relations are successively retrieved, the displayed page comprising at least one document characteristic key words is generated based on search result
Step comprising:
The second level mapping relations between document characteristic key words and event class standard statement word are established, standardization statement is saved in
Library；
Based on the second level mapping relations between document characteristic key words and event class standard statement word, the event class standard is traversed
It states all event class standards that lexon is concentrated and states word, obtain the corresponding text of each event class standard statement word
Book characteristic key words；
Statistical disposition is carried out to the document characteristic key words, generates the displaying comprising at least one document characteristic key words
The page.
9. according to the method described in claim 8, wherein, it is described establish document characteristic key words and event class standard statement word it
Between second level mapping relations the step of include:
Keyword extraction is carried out to all documents in document database and obtains document characteristic key words；
One-to-one or many-to-one second level mapping between the document characteristic key words and event class standard statement word is established to close
System.
10. a kind of searching system based on biaxial stress structure, which is characterized in that the system comprises:
Keyword-extraction module (10), the event information for obtaining to client carry out keyword extraction, obtain comprising at least
The set of one event non-standard Feature Words；
Standard scale predicate obtains module (20), for based on one between event non-standard Feature Words and event class standard statement word
Grade mapping relations, it is non-to obtain each event in the event non-standard feature set of words to the keyword-extraction module (10)
Specification features word is standardized, to obtain at least one corresponding event class standard table of each event non-standard Feature Words
Predicate forms event class standard and states set of words；
Membership analysis module (30), for traversing event class standard table all in the event class standard statement set of words
Predicate is stated the rank membership between word according to event class standard and is successively opened up to the event class standard statement word in the set
Membership analysis two-by-two is opened, to obtain the subset of event class standard statement set of words；
Search result generation module (40), for being reflected based on the second level between document characteristic key words and event class standard statement word
Relationship is penetrated, second level is successively retrieved to each of the event category fiducial mark statement lexon concentration event class standard statement word and is reflected
Relationship is penetrated, the displayed page comprising at least one document characteristic key words is generated based on search result.
11. system according to claim 10, wherein the system also includes: event information obtains module (50), is used for
The retrieval log in nearest setting time length is obtained from search engine server and the retrieval log that will acquire is as thing
Part information；Or the event information for inputting this user in client is as event information.
12. the described in any item systems of 0-11 according to claim 1, wherein the standard scale predicate obtains module (20) and includes:
Unit (201) are established in level-one mapping, one for establishing between event non-standard Feature Words and event class standard statement word
Grade mapping relations are saved in standardization statement library；
Non-standard Feature Words Traversal Unit (202), for traversing event non-standard feature set of words { FnIn all non-standard it is special
Word is levied, each event non-standard Feature Words F is obtainednAt least one corresponding event class standard states word Sm, formed comprising more
The event class standard of a event class standard statement word states set of words { Sm, n >=m >=1.
13. system according to claim 12, wherein the Feature Words Traversal Unit (202) further comprises:
Approximate statement word acquiring unit (2020), for being stated in library not when standard scale predicate Traversal Unit (202) from standardization
Obtain some described event non-standard Feature Words FiCorresponding event class standard states word Si, pass through Text similarity computing method
Standardization statement library is filtered, is obtained and event non-standard Feature Words FiApproximate event class standard states word Sj, 1≤i≤
N, 1≤j≤n；
Level-one map updating unit (2021), for establishing event non-standard Feature Words FiWith the approximate event class standard table
Predicate SjBetween level-one mapping relations, and the event non-standard Feature Words and the approximate event class standard are stated into word SjIt
Between level-one mapping relations be saved in standardization statement library in.
14. system according to claim 13, wherein the level-one mapping establishes unit (201) and further comprises:
User behavior analysis unit (2010), the user behavior data got for capturing client therefrom extract and obtain institute
State at least one event non-standard Feature Words；
Unit (201) are established in the level-one mapping, for based at least one described event non-standard Feature Words, foundation to be described extremely
One-to-one, one-to-many or many-to-one level-one between few event non-standard Feature Words and event class standard statement word
Mapping relations；
Wherein, the event non-standard Feature Words are the feature under event class standard statement word classification.
15. system described in 0,11 or 13 according to claim 1, wherein when the membership analysis module (30) detects
There are an event class standards to state word S in the event class standard statement set of wordspOr multiple identical event class standard tables
Predicate SpWhen, it is determined that the event class standard states word SpSubset as event class standard statement set of words.
16. system described in 0,11 or 13 according to claim 1, wherein when the membership analysis module (30) detects
When stating word there are multiple and different event class standards in the event class standard statement set of words, then to the event in the set
Membership analysis two-by-two is successively unfolded in class standard statement word, and the membership analysis module (30) specifically executes following behaviour
Make:
If event class standard states word SjWord S is stated with event class standardj+1Belong to coordination at the same level, it is determined that { Sj,Sj+1Be
The subset of event class standard statement set of words；
If event class standard states word SjWord S is stated with event class standardj+1Belong to covering relation at the same level, and event class standard is stated
Word SjCovering event class standard states word Sj+1, it is determined that { SjIt is the subset that the event class standard states set of words；
If event class standard states word SjWord S is stated comprising event class standardj+1, then { Sj+1It is that the event class standard states word set
The subset of conjunction；
If event class standard states word SjIt is contained in event class standard statement word Sj+1, then { SjIt is that the event class standard states word set
The subset of conjunction.
17. system described in 0,11 or 13 according to claim 1, wherein the search result generation module (40) includes:
Unit (401) are established in second level mapping, the second level for establishing between document characteristic key words and event class standard statement word
Mapping relations are saved in standardization statement library；
Standard scale predicate Traversal Unit (402), for based on two between document characteristic key words and event class standard statement word
Grade mapping relations traverse all event class standard statement words that the event class standard statement lexon is concentrated, obtain every
A event class standard states the corresponding document characteristic key words of word；
Displayed page generation unit (403) is generated comprising described in extremely for carrying out statistical disposition to the document characteristic key words
The displayed page of few document characteristic key words.
18. system according to claim 17, wherein the second level mapping establishes unit (401) and includes:
Document keyword extracting unit (4010) obtains text for carrying out keyword extraction to all documents in document database
Book characteristic key words；
Unit (401) are established in second level mapping, for establish the document characteristic key words and event class standard statement word it
Between one-to-one or many-to-one second level mapping relations.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610158099.7A CN105824938B (en) | 2016-03-18 | 2016-03-18 | A kind of search method and system based on biaxial stress structure |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610158099.7A CN105824938B (en) | 2016-03-18 | 2016-03-18 | A kind of search method and system based on biaxial stress structure |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105824938A CN105824938A (en) | 2016-08-03 |
| CN105824938B true CN105824938B (en) | 2019-11-08 |
Family
ID=56525215
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610158099.7A Active CN105824938B (en) | 2016-03-18 | 2016-03-18 | A kind of search method and system based on biaxial stress structure |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105824938B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109597986A (en) * | 2018-10-16 | 2019-04-09 | 深圳壹账通智能科技有限公司 | Localization method, device, equipment and the storage medium of abnormal problem |
| CN112802569B (en) * | 2021-02-05 | 2023-08-08 | 北京嘉和海森健康科技有限公司 | Semantic information acquisition method, device, equipment and readable storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102479230A (en) * | 2010-11-29 | 2012-05-30 | 北京四维图新科技股份有限公司 | Method and device for extracting geographic feature words |
| CN102760140A (en) * | 2011-04-29 | 2012-10-31 | 淮海工学院 | Incident body-based method for expanding searches |
| CN103116574A (en) * | 2013-02-22 | 2013-05-22 | 电子科技大学 | Method for mining domain process ontology from natural language text |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP4247284B2 (en) * | 2007-03-28 | 2009-04-02 | 株式会社東芝 | Information search apparatus, information search method, and information search program |
| KR20100046586A (en) * | 2008-10-27 | 2010-05-07 | 삼성전자주식회사 | Map-based web search method and apparatus |
-
2016
- 2016-03-18 CN CN201610158099.7A patent/CN105824938B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102479230A (en) * | 2010-11-29 | 2012-05-30 | 北京四维图新科技股份有限公司 | Method and device for extracting geographic feature words |
| CN102760140A (en) * | 2011-04-29 | 2012-10-31 | 淮海工学院 | Incident body-based method for expanding searches |
| CN103116574A (en) * | 2013-02-22 | 2013-05-22 | 电子科技大学 | Method for mining domain process ontology from natural language text |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105824938A (en) | 2016-08-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Luo et al. | Newsclippings: Automatic generation of out-of-context multimodal media | |
| US11816078B2 (en) | Automatic entity resolution with rules detection and generation system | |
| Motoyama et al. | I seek you: searching and matching individuals in social networks | |
| US9495445B2 (en) | Document sorting system, document sorting method, and document sorting program | |
| US20110125746A1 (en) | Dynamic machine assisted informatics | |
| Hutchins et al. | Hiding in plain sight: criminal network analysis | |
| Joshi et al. | Political issue extraction model: A novel hierarchical topic model that uses tweets by political and non-political authors | |
| CN111400448A (en) | Method and device for analyzing incidence relation of objects | |
| CN105824938B (en) | A kind of search method and system based on biaxial stress structure | |
| CN107544961A (en) | A kind of sentiment analysis method, equipment and its storage device of social media comment | |
| Bródka et al. | Profile cloning detection in social networks | |
| Jiang et al. | Camera fingerprint: A new perspective for identifying user's identity | |
| JP7092194B2 (en) | Information processing equipment, judgment method, and program | |
| CN107943935B (en) | Data processing method and device and computer readable storage medium | |
| Varol et al. | The importance of web activities for computer forensics | |
| Hu et al. | An empirical study on the names of points of interest and their changes with geographic distance | |
| US20150201104A1 (en) | Three-dimensional image searching based on inputs collected by a mobile device | |
| CN112907308B (en) | Data detection method and device, and computer readable storage medium | |
| Valldor et al. | Firearm detection in social media images | |
| CN108073567A (en) | A kind of Feature Words extraction process method, system and server | |
| CN105786929A (en) | Information monitoring method and device | |
| CN107133644B (en) | Digital library's content analysis system and method | |
| WO2019187107A1 (en) | Information processing device, control method, and program | |
| Wang et al. | Modeling sentiment evolution for social incidents | |
| Lapatta | Ecotourism recommendations based on sentiments using skyline query and apache-spark |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |