CN101770482A - Method and system for issuing advertisements - Google Patents
Method and system for issuing advertisements Download PDFInfo
- Publication number
- CN101770482A CN101770482A CN200810246860A CN200810246860A CN101770482A CN 101770482 A CN101770482 A CN 101770482A CN 200810246860 A CN200810246860 A CN 200810246860A CN 200810246860 A CN200810246860 A CN 200810246860A CN 101770482 A CN101770482 A CN 101770482A
- Authority
- CN
- China
- Prior art keywords
- advertisement
- keyword
- page
- clicking rate
- matched
- 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.)
- Pending
Links
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Information Transfer Between Computers (AREA)
Abstract
The invention discloses a method for issuing advertisements, comprising the following steps: carrying out statistics on click rate when the advertisement and keywords which can represent the interface are simultaneously demonstrated; when needing to issue advertisements to an interface to be matched, extracting the keywords which can represent the interface to be matched; learning about the matched advertisement which can represent the keywords of the interface to be matched according to the click rate when the advertisement and keywords which can represent the interface are simultaneously demonstrated; issuing the matched advertisement to the interface to be matched. The invention also discloses a system for issuing advertisements. With the method and the system of the invention, the potential link between the advertisements and interfaces can be found for instructing issuance of advertisements; as a result, interface advertisement issuance can be accurate, accurate matching between interfaces and advertisements can be realized to improve advertising effect effectively.
Description
Technical field
The present invention relates to the browser technology field, particularly relate to a kind of method and system of advertisement putting.
Background technology
Large-scale website is (such as portal website, the perhaps online special so vertical website of specialty that is skillful in a certain field, Zhong Guan-cun) has a lot of customer groups usually, and this part user's mass ratio is higher, or is that purchasing power is relatively arranged, or strong tendency is arranged.Therefore, some advertisers can require to these website advertisement deliveries, to reach the purpose of promoting product or expansion company popularity etc.
Purpose by the website showing advertisements is that the guiding user pays close attention to advertisement content, if thrown in advertisement on certain page in certain website, but the most of user who browses this page does not pay close attention to this advertisement, will can not embody the value of advertisement so, can't reach advertising effect.Therefore, need and carry out advertisement putting targetedly, to promote advertising results.
Because large-scale website utilizes channel to organize different articles usually, the article of same channel has identical focus usually.Based on this, can come advertisement delivery in the following manner at present: after getting access to the advertisement that advertiser provides, specify the channel of throwing in, in representing this channel, represent this advertisement then on the page of each piece article in some way according to the kind of advertisement.For example, with the advertisement putting of sports goods at sports channel, or the like.
But this method to a certain class page advertisement delivery, may not be on all pages, all to obtain good advertising results, for example, the advertisement putting of certain sports goods is arrived each page of sports channel, but certain page of sports channel is introduced sports lottery ticket, and then this advertisement may can't cause the user's who browses this page concern.
Summary of the invention
In view of this, the object of the present invention is to provide a kind of method and system of advertisement putting, can realize being accurate to the advertisement putting of the page.
For achieving the above object, the invention provides following scheme:
A kind of method of advertisement putting comprises:
Clicking rate when adding up advertisement in advance and representing simultaneously with keyword that can representing pages;
When needs during to certain page advertisement delivery to be matched, extraction can be represented the keyword of the described page to be matched; Clicking rate when representing simultaneously according to the advertisement of described statistics in advance and keyword that can representing pages is known the advertisement that is complementary with the described keyword that can represent the described page to be matched;
The described page to be matched is arrived in the described advertisement putting that is complementary.
Preferably, also comprise:
On the described page to be matched, represent the described advertisement that is complementary.
Preferably, describedly on the described page to be matched, represent the described advertisement that is complementary and comprise:
Clicking rate when representing simultaneously according to the described advertisement that is complementary and the described keyword that can represent the page to be matched is calculated the expectation clicking rate of advertisement at the described page to be matched;
Size according to described expectation clicking rate represents the described advertisement that is complementary on the described page to be matched.
Preferably, describedly on the described page to be matched, represent the described advertisement that is complementary according to described expectation clicking rate and comprise:
According to the size of described expectation clicking rate, the advertisement that is complementary of preset number is represented on the described page to be matched in turn;
Perhaps, utilizing described expectation clicking rate is the advertisement allocation probability weight that is complementary, and represents the advertisement that is complementary according to described probability right.
Preferably, count the clicking rate of advertisement when representing simultaneously with keyword that can representing pages after, also comprise:
Clicking rate when the average clicking rate of utilizing advertisement represents simultaneously to the advertisement that counts and keyword that can representing pages is done and is fallen power;
And/or,
The total degree that utilizes advertisement and keyword that can representing pages to represent simultaneously, the clicking rate the when advertisement that counts and keyword that can representing pages are represented is simultaneously done weighting.
Clicking rate when preferably, the keyword of described statistics advertisement and representing pages represents simultaneously comprises:
With the dimension coordinate of advertisement as matrix, keyword that can representing pages is as another dimension coordinate of matrix; Set up the clicking rate matrix of advertisement when representing simultaneously with keyword that can representing pages;
Described clicking rate when representing simultaneously according to described advertisement and keyword that can representing pages, obtain the advertisement that is complementary with the described keyword that can represent the page to be matched and comprise:
From described matrix, obtain the advertisement that is complementary with the described keyword that can represent the page to be matched.
A kind of system of advertisement putting comprises:
Statistic unit, the clicking rate when being used for adding up in advance advertisement and representing simultaneously with keyword that can representing pages;
Keyword extracting unit is used for when needs during to certain page advertisement delivery to be matched, and extraction can be represented the keyword of the described page to be matched;
The advertisement matching unit is used for the clicking rate when representing simultaneously according to the advertisement of described statistics in advance and keyword that can representing pages, knows the advertisement that is complementary with the described keyword that can represent the page to be matched;
The advertisement putting unit is used for the described advertisement putting that is complementary to the described page to be matched.
Preferably, also comprise:
The showing advertisement unit is used for representing the described advertisement that is complementary on the described page to be matched.
Preferably, described showing advertisement unit comprises:
Computation subunit, the clicking rate that is used for when representing simultaneously according to the described advertisement that is complementary and the described keyword that can represent the page to be matched is calculated the expectation clicking rate of advertisement at the described page to be matched;
Represent subelement, be used on the described page to be matched, representing the described advertisement that is complementary according to the size of described expectation clicking rate.
Preferably, the described size that represents subelement according to described expectation clicking rate represents the advertisement that is complementary of preset number in turn on the described page to be matched; Perhaps, utilizing described expectation clicking rate is the advertisement allocation probability weight that is complementary, and represents the advertisement that is complementary according to described probability right.
Preferably, also comprise:
Fall the power unit, the clicking rate when the average clicking rate that is used to utilize advertisement represents simultaneously to the advertisement that counts and keyword that can representing pages is done and is fallen power;
As/or,
Weighted units, the total degree that is used to utilize advertisement and keyword that can representing pages to represent simultaneously, the clicking rate the when advertisement that counts and keyword that can representing pages are represented is simultaneously done weighting.
Preferably, described statistic unit comprises:
Matrix is set up subelement, is used for the dimension coordinate of advertisement as matrix, and keyword that can representing pages is as another dimension coordinate of matrix; Set up the clicking rate matrix of advertisement when representing simultaneously with keyword that can representing pages;
Described advertisement matching unit obtains the advertisement that is complementary with the described keyword that can represent the page to be matched from described matrix.
According to specific embodiment provided by the invention, the invention discloses following technique effect:
Clicking rate when the present invention represents with keyword that can representing pages simultaneously by adding up advertisement in advance; Like this when needs during to certain page advertisement delivery to be matched, just can at first obtain the keyword that to represent the described page to be matched, clicking rate when representing simultaneously according to the advertisement that counts in advance and keyword that can representing pages then, obtain the advertisement that is complementary with the described keyword that can represent the page to be matched, and with the described advertisement putting that is complementary to the described page to be matched.As seen, the present invention is based on can representing pages keyword and the clicking rate of advertisement when representing simultaneously, having excavated exist between the advertisement and the page potential gets in touch, and instruct the input of advertisement with this, therefore can be as accurate as the input that the page carries out advertisement, realize the accurate coupling of the page and advertisement, thereby promoted advertising results effectively.
In addition, power falls in the clicking rate when the average clicking rate by advertisement represents simultaneously to the advertisement that counts and keyword that can representing pages, got rid of advertisement itself and made the influence of factor such as bad ad click rate, feasible advertisement of throwing in to certain page to be matched, be that clicking rate promotes the advertisement of big (with respect to representing) when representing on this page on other pages, rather than the bigger advertisement of clicking rate own.Simultaneously, can also utilize advertisement and the total degree that the keyword that can represent webpage represents simultaneously to do weighting (many more with the occurrence number, weighting is big more), can pick out more stable information by this method of weighting, the exclusive segment noise promotes the effect of throwing in.
Description of drawings
Fig. 1 is the process flow diagram of the method that provides of the embodiment of the invention;
Fig. 2 is another process flow diagram of the method that provides of the embodiment of the invention;
Fig. 3 is the synoptic diagram of the system that provides of the embodiment of the invention.
Embodiment
For above-mentioned purpose of the present invention, feature and advantage can be become apparent more, the present invention is further detailed explanation below in conjunction with the drawings and specific embodiments.
Referring to Fig. 1, the method for the advertisement putting that the embodiment of the invention provides may further comprise the steps:
S101: the clicking rate when adding up advertisement in advance and representing simultaneously with keyword that can representing pages;
Before carrying out concrete advertisement putting, the total degree that can be in advance represents simultaneously by statistics advertisement and keyword that can representing pages, and the clicked total degree of advertisement when representing simultaneously, calculate the clicking rate of advertisement when representing simultaneously with keyword that can representing pages.Wherein, advertisement may be a plurality of, and keyword that can representing pages also may be a plurality of.
S102: when needs during to certain page advertisement delivery to be matched, extraction can be represented the keyword of the described page;
S103: the clicking rate when representing simultaneously according to the advertisement of described statistics in advance and keyword that can representing pages, know the advertisement that is complementary with the described keyword that can represent the page to be matched;
S104: the described page to be matched is arrived in the described advertisement putting that is complementary.
In fact, exist some potential getting in touch between advertisement and the page info, excavate these potential contacts, make the advertisement of input on certain page embody its value veritably.For example, introduce the advertisement of throwing in certain brand automobile on the page of automobile market at one, the user who then browses this page has very high probability and clicks this advertisement.
Therefore, the core of the embodiment of the invention is, clicking rate when representing simultaneously by analyzing keyword that each advertisement and each can representing pages is judged certain page and is fit to throw in which advertisement, then with the advertisement putting that is fit to the corresponding page, promptly realize being accurate to the advertisement putting of the page, to improve advertising results.Be fit to throw in which advertisement in order to judge certain page, the embodiment of the invention has adopted analysis showing advertisement and click logs to obtain the method for effective information.
Wherein, having write down each advertisement in described showing advertisement and the click logs represented on which page, and represented on the page of certain advertisement at each, the clicking rate of this advertisement can be utilized these information to instruct and throw in which advertisement to the page to be matched.Below just the advertisement putting that how to utilize these information to instruct to be accurate to the page is at length introduced.
If will judge to the page to be matched to throw in which advertisement by the statistics clicking rate of each advertisement on each page, prerequisite be the page added up with the page to be matched between have that certain is related.On the other hand, the content that represents in the page all can comprise content of text usually, and can comprise the keyword that can represent this main contents of webpages usually in the content of text.Therefore, the present invention is exactly according to These characteristics, and the keyword that use can the representing pages content associates each page.In order to understand method provided by the invention better, the method for the advertisement delivery that the embodiment of the invention is provided below by concrete example describes in detail.Referring to Fig. 2, mainly may further comprise the steps in this example:
S201: when adding up keyword that each advertisement and each can representing pages and representing simultaneously, the clicking rate of advertisement.
In this step, can add up the click relevant information of advertisement, comprise whether clicked and clicking rate of the representing of advertisement and keyword that can representing pages, advertisement etc.Concrete, this step can comprise:
1) from advertisement click logs, extracts the content of text that the advertisement that certain page that the user browses represented and the page mainly represent, and whether the statistics user has clicked advertisement in navigation process, what click is which advertisement (can simultaneously or represent a plurality of advertisements in turn on a page sometimes), and has what users to click this advertisement.
Why needing to add up the whether number of times of clicked mistake and click of each advertisement, is that these advertisements may appear at the diverse location of the page simultaneously, also may represent in turn in identical position because may represent a plurality of advertisements on a page.Whether but the advertisement that represents on the same page may not necessarily all can both cause user's concern, therefore, merely add up each advertisement and represented on certain page, can not represent this advertisement can obtain good advertising results on this page.Therefore, what the present invention need add up is advertisement clicked mistake on which page once, and the clicking rate on certain page has much; And clicking rate is high more, illustrates that this page promotes bigger to this advertising effect.
2) content of text that the page is mainly represented is cut speech, and the speech that cuts out is carried out the entry mark.
Suppose that " Beijing is the capital of China to complete words.", cut speech and be meant each entry is all distinguished---" Beijing/be/China// capital." part-of-speech tagging is meant that part of speech (noun, verb, the adjective etc.) mark each speech comes out---" Beijing (noun)/be (link-verb)/China's (noun)/(auxiliary word)/capital (noun).”
3) keyword that screening can representing pages from the entry that cuts out.
Utilize conventional I R model (for example tf, idf model or BM2500 model), extract the keyword (may be a plurality of) that to represent this page.
Wherein, IR (information retrieval) model is meant a series of models that are applied to information retrieval, wherein:
Tf: the number of times that speech occurs in article;
Idf:df is meant a speech occurred, and idf is exactly the inverse of df in what piece articles;
The tf/idf model: this model thinks that speech occurrence number in this piece article is many more, and occurrence number is few more in other articles just can represent this piece article more.Citation form is similar to td*idf;
BM2500 model: relational noun qtf---(number of times that the speech number that speech is retrieved in search engine, speech are retrieved is many more, might represent one piece of article more), the BM2500 model is one and has used tf, idf, the computing formula of qtf.
By these models just can extract can representing pages keyword.But the keyword that extracts can not guarantee entirely accurate, therefore, in order to improve algorithm effect, can utilize statistical information+rule and method to filter out the noise speech.For example utilize to represent number of times (too much or very few) and blacklist filters out a part of noise speech (rule and method), re-use certain information retrieval model (for example click side etc.) and filter out a part of noise speech (statistical method).It not is necessary step that the noise speech filters, but the effect of the tangible boosting algorithm of filtering noise speech meeting.
Wherein, why to filter out representing too much entry, be because, if it is too much that entry represents number of times, illustrate that then the scope that this speech occurs is wider, may in the high page of " automotive-type, notebook class, milk class " these ad click rates, frequently occur, as a result this speech property distinguished not, can't know when this speech occurring which concrete advertisement this recommends, because the clicking rate of each advertisement of this speech correspondence is all similar.
Blacklist: such as ",,, you, I, he " like this can be in various articles the frequent speech that occurs, and " Beijing, Ministry of Foreign Affairs, dichloro-benzenes " is though thisly only occur in a part of article, but may not instruct the speech (because these speech do not have commercial tendency, " milk, IBM, motor vehicle purchase tax " this speech is only the speech of commercial tendency comparatively speaking) of advertisement putting.
4) utilize the keyword that obtains after the filtration, set up the clicking rate tables of data of the keyword of advertisement-representing pages, during specific implementation, this tables of data can be represented with the form of matrix: advertisement is a dimension coordinate of matrix, keyword is another dimension coordinate of matrix, a known advertisement and a keyword then can find the clicking rate of this advertisement when this keyword occurs in matrix.
Other each pages are also done same processing respectively, finally obtain the total degree S1 that keyword that each advertisement and each can representing pages represents simultaneously, and the clicked total degree S2 of each advertisement when representing simultaneously, and the clicking rate S2/S1 when calculating keyword that each advertisement and each can representing pages and representing simultaneously.
For example, in " the brand-new A4L of Audi listing ceremony on November 16 " this piece article " Audi, automobile, wheelbase, four-wheel drive " these keywords are arranged, it is respectively inferior for " 10000; 5000; 20000 " that simultaneously the user once represented " BMW, Mongolia Ox, with square computer " when browsing this piece article, each clicked " 100; 10; 5 " inferior, think that then above these keywords " Audi, automobile, wheelbase, four-wheel drive " and above advertisement " BMW " with having represented now 10000 times, clicked 100 times ....Add up and to obtain keyword " Audi, automobile, wheelbase, four-wheel drive " behind a plurality of pages and advertisement " BMW " is inferior with having represented " 200000,250000; 150000 " now, it is inferior to click " 200; 400; 500 ", be exactly " 0.1%, 0.16%, 0.333% " with current clicking rate so.
In actual applications, owing to reasons such as the good degree varies of making between the advertisement, can be from influencing the clicking rate of advertisement to a great extent.This just means that the advertisement that comes out and different keyword are with the difference between the current clicking rate, the factor that can have the good degree of advertisement itself, therefore need be with the advertisement that comes out and certain keyword with the clicking rate of now the clicking rate of each advertisement divided by showing advertisements at random the time, calculate the lifting ratio of ad click rate.Could really embody value like this to page A recommended advertisements a, otherwise, can tend to represent high advertisement of clicking rate own rather than the high advertisement of clicking rate lifting.
That is to say that the embodiment of the invention can also be done at the average clicking rate of advertisement and fall power (average clicking rate is high more, falls Quan Yueda) after counting advertisement and the keyword clicking rate with current advertisement.For example, if advertisement a clicking rate Sclick1 under random case is 0.1%, calculating clicking rate clickA1 when certain keyword A occurs is 0.18%, another one advertisement b clicking rate Sclick2 under random case is 0.05%, the clicking rate clickA2 that calculates when keyword A occurs is 0.15%, although clickA1＞clickA2 so, should be more prone to advertisement delivery b when but speech A obviously occurring, need at first to calculate clickA1/Sclick1 for this reason, clickA2/Sclick2, and then compare.
In addition, can also and can represent the same occurrence number of the keyword of webpage to do weighting (many more with the occurrence number, weighting is big more), can pick out more stable information by this method of weighting, the exclusive segment noise according to advertisement.Reason is many more with the occurrence number, and statistical information just can reflect truth more, otherwise then reflects more cas fortuits, and therefore, if do not use the method, the noise that the part accidentalia is brought just can not filter, and can reduce algorithm effects.
S202: the coupling of the page and advertisement.
The keyword of having added up advertisement and representing pages content among the step S201 is with current, and the click information of advertisement just can utilize the above-mentioned information that comes out to carry out at the advertisement putting of the page then, specifically can may further comprise the steps:
1) obtains the page (can when user capture, obtain the page that the user browses in real time) to be matched, the method of using when utilizing statistics is cut speech and is carried out part-of-speech tagging the content of text in the page, and same method extracts the keyword that can represent this content of pages when utilizing with statistics then.
2) from the clicking rate matrix of advertisement-keyword correspondence of having set up, each advertisement of obtaining each keyword correspondence that can represent this content of pages, and corresponding respectively clicking rate.
3) because keyword may more than one, and the advertisement of each keyword correspondence may also be a plurality of, and the advertisement of each keyword correspondence may have overlapping, therefore, can also the expectation clicking rate of each advertising words of COMPREHENSIVE CALCULATING on the page.Owing to got access to the clicking rate of each advertisement when each keyword speech occurs, as long as the clicking rate of therefore utilizing the geometric mean method to calculate each keyword that current page extracts just can obtain each advertisement in the expectation clicking rate of this page.
For example, the keyword that extracts has two, is respectively A, B, wherein the advertisement of A correspondence be a, b,, corresponding clicking rate is respectively 0.2,0.3; The advertisement of B correspondence also is a, b, and corresponding clicking rate is 0.5,0.1.Then calculating the expectation clicking rate of advertisement a on this page is
The expectation clicking rate of advertisement b on this page is
In the process of aforementioned calculation, can also utilize the weight of each keyword to do weighting (weight of keyword in this page is high more, and corresponding ad click rate weighting is big more).There emerged a the expectation clicking rate of advertisement in calculating after, the size of clicking rate sorts to each advertisement according to expectation.
S203: will on this page, represent with the advertisement that this page is complementary.
The described advertisement that each is complementary can be represented on this page to be matched at random, can also represent in proper order according to the advertisement that calculates in the above-mentioned steps, concrete ways of presentation can be: before choosing K advertisement according to equal proportion or according to expectation the big or small allocation proportion of clicking rate on this page, show in turn, for example, by estimating that clicking rate represents certain number of times from big to small in turn, at this moment, suppose the advertisement that is complementary according to expectation clicking rate sort from big to small and be A, B, C, when then representing, showing advertisements A for the first time, showing advertisements B for the second time, showing advertisements C for the third time, the 4th showing advertisements A...... again; Certainly, also can only show that ordering is the highest.In addition, can also be to estimate that clicking rate is a probability right, and represent the advertisement that respectively is complementary according to this probability right, for example: the advertisement that is complementary is A, B, suppose the weight weightA=1 of advertisement A, the weight weightB=2 of advertisement B is then during showing advertisements, showing advertisements A under 1/3 situation, showing advertisements B under 2/3 situation.
To sum up, the method that the embodiment of the invention adopts is exactly, at first set up each advertisement and can each keyword of representing pages between corresponding relation, like this, for certain page to be matched, as long as extract the keyword that to represent this page, just can be according to the corresponding relation between advertisement of having set up and the keyword, find the corresponding advertisement of the keyword of representing this page to be matched, and these showing advertisements are got final product on this page to be matched.Therefore, compare with the method for throwing in certain series advertisements at channel in the prior art, the coupling granularity is thinner, can be as accurate as the input that the page carries out advertisement, has realized the accurate coupling of the page and advertisement, can promote advertising results effectively.
In addition, it should be noted that the clicking rate the when embodiment of the invention represents simultaneously based on keyword that can representing pages and advertisement is carried out, rather than merely mate based on the keyword that occurs on the page.For example, " automobile " wording is arranged on the page,, then may directly match the advertisement of certain brand automobile if merely mate based on the keyword that occurs on the page.But in fact this page may be introduced the house, and only wherein automobile has been mentioned in the somewhere, and therefore at this moment going out automotive advertising might not get good result; Based on this, the embodiment of the invention can at first extract can representing pages keyword, and then the clicking rate when representing simultaneously according to advertisement that counts and keyword that can representing pages is known the advertisement that is complementary with this page.Because the above-mentioned page is mainly introduced the house, automobile be not can this page keyword, therefore, the advertisement that final that obtain and this page are complementary will be relevant with the house (these advertisements might not comprise " house " two words, but these advertisements are the advertisement that has higher clicking rate when representing simultaneously with keyword " house " certainly), therefore, the embodiment of the invention helps promoting advertising effect.
The method of the advertisement putting that provides with the embodiment of the invention is corresponding, and the embodiment of the invention also provides a kind of system of advertisement putting, and referring to Fig. 3, this system comprises with lower unit:
Statistic unit U301, the clicking rate when being used for adding up in advance advertisement and representing simultaneously with keyword that can representing pages;
Keyword extracting unit U302 is used for when needs during to certain page advertisement delivery to be matched, and extraction can be represented the keyword of the described page to be matched;
Advertisement matching unit U303 is used for the clicking rate when representing simultaneously according to the advertisement that counts and keyword that can representing pages, knows the advertisement that is complementary with the described keyword that can represent the described page to be matched;
Advertisement putting unit U304 is used for the described advertisement putting that is complementary to the described page to be matched.
Clicking rate when statistic unit U301 adds up advertisement in advance and represents simultaneously with keyword that can representing pages; The total degree that can represent simultaneously by statistics advertisement and keyword that can representing pages, and the clicked total degree of advertisement calculates described clicking rate when representing simultaneously, statistics just can be accurate to the advertisement putting of the page later like this, in concrete launch process, at first extract the keyword that to represent the page to be matched by keyword extracting unit U302; Clicking rate when being represented simultaneously according to the described advertisement that come out and keyword that can representing pages by advertisement matching unit U303 is then obtained the advertisement that is complementary with the described keyword that can represent the page to be matched; At last, advertisement putting unit U304 gets final product the described advertisement putting that is complementary to the described page to be matched.
Determining after which advertisement certain page to be matched throws in, just can on the page to be matched, carry out having represented of advertisement, therefore, this system can also comprise:
Showing advertisement unit U305 is used for representing the described advertisement that is complementary on the described page to be matched.
Wherein, concrete ways of presentation can be arbitrarily, also can represent according to pre-specified ways of presentation.For example, showing advertisement unit U305 can comprise:
Computation subunit U3051, the clicking rate that is used for when representing simultaneously according to the described advertisement that is complementary and the described keyword that can represent the page to be matched is calculated the expectation clicking rate of each advertisement at the described page to be matched;
Represent subelement U3052, be used on the described page to be matched, representing the described advertisement that is complementary according to the size of described expectation clicking rate.
Wherein, representing subelement U3052 can represent the size of each advertisement according to described expectation clicking rate in turn on the described page to be matched; Also can on the described page to be matched, represent the advertisement of the preset number of estimating that clicking rate is the highest.
In order further to promote effect of the present invention, this system can also comprise:
Fall power unit U306, the clicking rate the when advertisement that the average clicking rate that is used to utilize advertisement counts statistic unit U301 represents simultaneously with keyword that can representing pages is done and is fallen power.Utilize this to fall power unit U306 and can calculate the lifting ratio of ad click rate on certain page, thereby eliminate the influence of factors such as the good degree of advertisement itself.Could really embody value like this to page A recommended advertisements a, otherwise, can tend to represent high advertisement of clicking rate own rather than clicking rate and on this page, promote high advertisement.
Can also comprise weighted units U307, the total degree that is used to utilize advertisement and keyword that can representing pages to represent simultaneously, the clicking rate the when advertisement that statistic unit U301 is counted represents simultaneously with keyword that can representing pages is done weighting.By weighted units U307, can pick out more stable information, exclusive segment noise, the effect of the lifting embodiment of the invention.
After counting the clicking rate of advertisement when representing simultaneously with keyword that can representing pages, be equivalent to obtain following information: advertisement, with advertisement represent simultaneously can representing pages keyword and advertisement and keyword that can the representing pages clicking rate when representing simultaneously, for the ease of inquiry, can utilize the form of matrix to preserve these information.Therefore, statistic unit can comprise:
Matrix is set up subelement U3011, is used for the dimension coordinate of advertisement as matrix, and keyword that can representing pages is as another dimension coordinate of matrix; Set up the clicking rate matrix of advertisement when representing simultaneously with keyword that can representing pages; Like this, a known advertisement and a keyword, advertisement matching unit U303 then can find the clicking rate of this advertisement when this keyword occurs easily in matrix, and then obtains each advertisement that is complementary with described each keyword that can represent the page to be matched.
More than to the method and system of a kind of advertisement putting provided by the present invention, be described in detail, used specific case herein principle of the present invention and embodiment are set forth, the explanation of above embodiment just is used for helping to understand method of the present invention and core concept thereof; Simultaneously, for one of ordinary skill in the art, according to thought of the present invention, part in specific embodiments and applications all can change.In sum, this description should not be construed as limitation of the present invention.
Claims (12)
1. the method for an advertisement putting is characterized in that, comprising:
Clicking rate when adding up advertisement in advance and representing simultaneously with keyword that can representing pages;
When needs during to certain page advertisement delivery to be matched, extraction can be represented the keyword of the described page to be matched; Clicking rate when representing simultaneously according to the advertisement of described statistics in advance and keyword that can representing pages is known the advertisement that is complementary with the described keyword that can represent the described page to be matched;
The described page to be matched is arrived in the described advertisement putting that is complementary.
2. method according to claim 1 is characterized in that, also comprises:
On the described page to be matched, represent the described advertisement that is complementary.
3. method according to claim 2 is characterized in that, describedly represents the described advertisement that is complementary comprise on the described page to be matched:
Clicking rate when representing simultaneously according to described each advertisement that is complementary and described each keyword that can represent the page to be matched is calculated the expectation clicking rate of each advertisement at the described page to be matched;
Size according to described expectation clicking rate represents the described advertisement that is complementary on the described page to be matched.
4. method according to claim 3 is characterized in that, describedly represents the described advertisement that is complementary according to described expectation clicking rate on the described page to be matched and comprises:
According to the size of described expectation clicking rate, the advertisement that is complementary of preset number is represented on the described page to be matched in turn;
Perhaps, utilizing described expectation clicking rate is the advertisement allocation probability weight that is complementary, and represents the advertisement that is complementary according to described probability right.
5. method according to claim 1 is characterized in that, count the clicking rate of advertisement when representing simultaneously with keyword that can representing pages after, also comprise:
Clicking rate when the average clicking rate of utilizing advertisement represents simultaneously to the advertisement that counts and keyword that can representing pages is done and is fallen power;
And/or,
The total degree that utilizes advertisement and keyword that can representing pages to represent simultaneously, the clicking rate the when advertisement that counts and keyword that can representing pages are represented is simultaneously done weighting.
6. method according to claim 1 is characterized in that, the clicking rate the when keyword of described statistics advertisement and representing pages represents simultaneously comprises:
With the dimension coordinate of advertisement as matrix, keyword that can representing pages is as another dimension coordinate of matrix; Set up the clicking rate matrix of advertisement when representing simultaneously with keyword that can representing pages;
Described clicking rate when representing simultaneously according to described advertisement and keyword that can representing pages, obtain the advertisement that is complementary with the described keyword that can represent the page to be matched and comprise:
From described matrix, obtain the advertisement that is complementary with the described keyword that can represent the page to be matched.
7. the system of an advertisement putting is characterized in that, comprising:
Statistic unit, the clicking rate when being used for adding up in advance advertisement and representing simultaneously with keyword that can representing pages;
Keyword extracting unit is used for when needs during to certain page advertisement delivery to be matched, and extraction can be represented the keyword of the described page to be matched;
The advertisement matching unit is used for the clicking rate when representing simultaneously according to the advertisement of described statistics in advance and keyword that can representing pages, knows the advertisement that is complementary with the described keyword that can represent the page to be matched;
The advertisement putting unit is used for the described advertisement putting that is complementary to the described page to be matched.
8. system according to claim 7 is characterized in that, also comprises:
The showing advertisement unit is used for representing the described advertisement that is complementary on the described page to be matched.
9. system according to claim 8 is characterized in that, described showing advertisement unit comprises:
Computation subunit, the clicking rate that is used for when representing simultaneously according to the described advertisement that is complementary and the described keyword that can represent the page to be matched is calculated the expectation clicking rate of advertisement at the described page to be matched;
Represent subelement, be used on the described page to be matched, representing the described advertisement that is complementary according to the size of described expectation clicking rate.
10. system according to claim 9 is characterized in that, the described size that represents subelement according to described expectation clicking rate represents the advertisement that is complementary of preset number in turn on the described page to be matched; Perhaps, utilizing described expectation clicking rate is the advertisement allocation probability weight that is complementary, and represents the advertisement that is complementary according to described probability right.
11. system according to claim 7 is characterized in that, also comprises:
Fall the power unit, the clicking rate when the average clicking rate that is used to utilize advertisement represents simultaneously to the advertisement that counts and keyword that can representing pages is done and is fallen power;
And/or,
Weighted units, the total degree that is used to utilize advertisement and keyword that can representing pages to represent simultaneously, the clicking rate the when advertisement that counts and keyword that can representing pages are represented is simultaneously done weighting.
12. system according to claim 7 is characterized in that, described statistic unit comprises:
Matrix is set up subelement, is used for the dimension coordinate of advertisement as matrix, and keyword that can representing pages is as another dimension coordinate of matrix; Set up the clicking rate matrix of advertisement when representing simultaneously with keyword that can representing pages;
Described advertisement matching unit obtains the advertisement that is complementary with the described keyword that can represent the page to be matched from described matrix.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200810246860A CN101770482A (en) | 2008-12-26 | 2008-12-26 | Method and system for issuing advertisements |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200810246860A CN101770482A (en) | 2008-12-26 | 2008-12-26 | Method and system for issuing advertisements |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101770482A true CN101770482A (en) | 2010-07-07 |
Family
ID=42503347
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN200810246860A Pending CN101770482A (en) | 2008-12-26 | 2008-12-26 | Method and system for issuing advertisements |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101770482A (en) |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102323951A (en) * | 2011-09-08 | 2012-01-18 | 中兴通讯股份有限公司 | Internet advertising method and system |
| CN102446171A (en) * | 2010-09-30 | 2012-05-09 | 百度在线网络技术（北京）有限公司 | Method and apparatus for evaluating quality score of promotion key word based on weighted average click-through rate |
| CN102467518A (en) * | 2010-11-05 | 2012-05-23 | 百度在线网络技术（北京）有限公司 | Method and equipment for setting necessary semantic components in popularization keyword |
| CN102646248A (en) * | 2012-02-27 | 2012-08-22 | 沈文策 | Advertisement publishing method and system |
| CN102760142A (en) * | 2011-04-29 | 2012-10-31 | 北京百度网讯科技有限公司 | Method and device for extracting subject label in search result aiming at searching query |
| CN102760138A (en) * | 2011-04-26 | 2012-10-31 | 北京百度网讯科技有限公司 | Classification method and device for user network behaviors and search method and device for user network behaviors |
| CN103093367A (en) * | 2011-10-28 | 2013-05-08 | 聚胜万合信息技术（上海）有限公司 | Precise internet advertisement putting system and method |
| CN103425680A (en) * | 2012-05-18 | 2013-12-04 | 阿里巴巴集团控股有限公司 | Selection method and system for page advertisement demonstration |
| CN104965893A (en) * | 2015-06-18 | 2015-10-07 | 山东师范大学 | Big data advertisement delivery method |
| CN106776665A (en) * | 2015-11-25 | 2017-05-31 | 北京国双科技有限公司 | The put-on method and device of keyword |
| CN107516247A (en) * | 2017-08-28 | 2017-12-26 | 天脉聚源（北京）科技有限公司 | A kind of method and device for predicting advertisement played data |
| WO2018014269A1 (en) * | 2016-07-20 | 2018-01-25 | 深圳市东信时代信息技术有限公司 | Optimization method and device for media advertisement click rate, and media advertisement system |
| CN110930969A (en) * | 2019-10-14 | 2020-03-27 | 科大讯飞股份有限公司 | Background music determination method and related equipment |
| CN111160953A (en) * | 2019-12-12 | 2020-05-15 | 浙江大华技术股份有限公司 | Method and device for delivering advertisement in application and storage medium |
| CN111429188A (en) * | 2020-04-09 | 2020-07-17 | 湘潭大学 | Webpage advertisement putting method |
| CN111723260A (en) * | 2019-03-19 | 2020-09-29 | 百度在线网络技术（北京）有限公司 | Method and device for acquiring recommended content, electronic equipment and readable storage medium |
-
2008
- 2008-12-26 CN CN200810246860A patent/CN101770482A/en active Pending
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102446171A (en) * | 2010-09-30 | 2012-05-09 | 百度在线网络技术（北京）有限公司 | Method and apparatus for evaluating quality score of promotion key word based on weighted average click-through rate |
| CN102467518A (en) * | 2010-11-05 | 2012-05-23 | 百度在线网络技术（北京）有限公司 | Method and equipment for setting necessary semantic components in popularization keyword |
| CN102760138A (en) * | 2011-04-26 | 2012-10-31 | 北京百度网讯科技有限公司 | Classification method and device for user network behaviors and search method and device for user network behaviors |
| CN102760142A (en) * | 2011-04-29 | 2012-10-31 | 北京百度网讯科技有限公司 | Method and device for extracting subject label in search result aiming at searching query |
| CN102323951A (en) * | 2011-09-08 | 2012-01-18 | 中兴通讯股份有限公司 | Internet advertising method and system |
| CN103093367A (en) * | 2011-10-28 | 2013-05-08 | 聚胜万合信息技术（上海）有限公司 | Precise internet advertisement putting system and method |
| CN102646248B (en) * | 2012-02-27 | 2016-01-13 | 沈文策 | A kind of advertisement delivery method and system |
| CN102646248A (en) * | 2012-02-27 | 2012-08-22 | 沈文策 | Advertisement publishing method and system |
| CN103425680A (en) * | 2012-05-18 | 2013-12-04 | 阿里巴巴集团控股有限公司 | Selection method and system for page advertisement demonstration |
| CN104965893A (en) * | 2015-06-18 | 2015-10-07 | 山东师范大学 | Big data advertisement delivery method |
| CN106776665A (en) * | 2015-11-25 | 2017-05-31 | 北京国双科技有限公司 | The put-on method and device of keyword |
| WO2018014269A1 (en) * | 2016-07-20 | 2018-01-25 | 深圳市东信时代信息技术有限公司 | Optimization method and device for media advertisement click rate, and media advertisement system |
| CN107516247A (en) * | 2017-08-28 | 2017-12-26 | 天脉聚源（北京）科技有限公司 | A kind of method and device for predicting advertisement played data |
| CN111723260A (en) * | 2019-03-19 | 2020-09-29 | 百度在线网络技术（北京）有限公司 | Method and device for acquiring recommended content, electronic equipment and readable storage medium |
| CN110930969A (en) * | 2019-10-14 | 2020-03-27 | 科大讯飞股份有限公司 | Background music determination method and related equipment |
| CN110930969B (en) * | 2019-10-14 | 2024-02-13 | 科大讯飞股份有限公司 | Background music determining method and related equipment |
| CN111160953A (en) * | 2019-12-12 | 2020-05-15 | 浙江大华技术股份有限公司 | Method and device for delivering advertisement in application and storage medium |
| CN111160953B (en) * | 2019-12-12 | 2023-11-03 | 浙江大华技术股份有限公司 | Method, device and storage medium for putting advertisement in application |
| CN111429188A (en) * | 2020-04-09 | 2020-07-17 | 湘潭大学 | Webpage advertisement putting method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101770482A (en) | Method and system for issuing advertisements | |
| US11176142B2 (en) | Method of data query based on evaluation and device | |
| Cheng et al. | Personalized click prediction in sponsored search | |
| JP5507154B2 (en) | System and method for expanding advertising targets based on topics using social indexing | |
| JP5456412B2 (en) | Community-based advertising word ambiguity removal system and method | |
| CN101436186B (en) | Method and system for providing related searches | |
| US20160162582A1 (en) | Method and system for conducting an opinion search engine and a display thereof | |
| CN100570611C (en) | A kind of methods of marking of the information retrieval document based on viewpoint searching | |
| CN103530299B (en) | Search result generating method and device | |
| US20100030647A1 (en) | Advertisement selection for internet search and content pages | |
| CN103606097A (en) | Method and system based on credibility evaluation for product information recommendation | |
| US20120197732A1 (en) | Action-aware intent-based behavior targeting | |
| CN103729359A (en) | Method and system for recommending search terms | |
| CN102841946A (en) | Commodity data retrieval sequencing and commodity recommendation method and system | |
| CN103729360A (en) | Interest label recommendation method and system | |
| CN103365839A (en) | Recommendation search method and device for search engines | |
| CN102982153A (en) | Information retrieval method and device | |
| US20160232247A1 (en) | Computerized internet search system and method | |
| CN106062743A (en) | Systems and methods for keyword suggestion | |
| CN106296368A (en) | A kind of vehicle commending system and method | |
| CN104424308A (en) | Web page classification standard acquisition method and device and web page classification method and device | |
| US20140046756A1 (en) | Generative model for related searches and advertising keywords | |
| CN103198118A (en) | Method and system for backtracking product web pages | |
| CN103425680A (en) | Selection method and system for page advertisement demonstration | |
| CN103778122A (en) | Searching method and system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C12 | Rejection of a patent application after its publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20100707 |