CN105956161B - A kind of information recommendation method and device - Google Patents

A kind of information recommendation method and device Download PDF

Info

Publication number
CN105956161B
CN105956161B CN201610327707.2A CN201610327707A CN105956161B CN 105956161 B CN105956161 B CN 105956161B CN 201610327707 A CN201610327707 A CN 201610327707A CN 105956161 B CN105956161 B CN 105956161B
Authority
CN
China
Prior art keywords
website information
information
search term
weight
under
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.)
Expired - Fee Related
Application number
CN201610327707.2A
Other languages
Chinese (zh)
Other versions
CN105956161A (en
Inventor
向园
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Qihoo Technology Co Ltd
Original Assignee
Beijing Qihoo Technology Co Ltd
Qizhi Software Beijing Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Beijing Qihoo Technology Co Ltd, Qizhi Software Beijing Co Ltd filed Critical Beijing Qihoo Technology Co Ltd
Priority to CN201610327707.2A priority Critical patent/CN105956161B/en
Publication of CN105956161A publication Critical patent/CN105956161A/en
Application granted granted Critical
Publication of CN105956161B publication Critical patent/CN105956161B/en
Expired - Fee Related legal-status Critical Current
Anticipated expiration legal-status Critical

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
    • G06F16/953—Querying, e.g. by the use of web search engines
    • G06F16/9535—Search customisation based on user profiles and personalisation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/02—Marketing; Price estimation or determination; Fundraising
    • G06Q30/0241—Advertisements
    • G06Q30/0251—Targeted advertisements
    • G06Q30/0255—Targeted advertisements based on user history

Landscapes

  • Engineering & Computer Science (AREA)
  • Databases & Information Systems (AREA)
  • Theoretical Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Finance (AREA)
  • Strategic Management (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Accounting & Taxation (AREA)
  • Development Economics (AREA)
  • General Engineering & Computer Science (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Game Theory and Decision Science (AREA)
  • Data Mining & Analysis (AREA)
  • Economics (AREA)
  • Marketing (AREA)
  • General Business, Economics & Management (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • Information Transfer Between Computers (AREA)

Abstract

The embodiment of the invention provides a kind of information recommendation methods, are related to ad system technical field.The described method includes: counting the weight of each website information under each search term according to the search behavior historical record of each user；According to the browsing behavior historical record of each user, the number of clicks of each website information under each user identifier is counted；According to the weight of each website information under the number of clicks and each search term of each website information under each user identifier, the weight of each search term of each website information under each user identifier is calculated；When user newly accesses website information, user identifier and website information are obtained, and recommendation information is generated according to the biggish search term of weight and is shown.The problem of it is not the network address accessed by search behavior for user that the present invention, which is solved, and ad system can not recommend accurate information, achieves and improves the accuracy that user directly accesses recommendation information when under network address, reduce the effect of system resource waste.

Description

A kind of information recommendation method and device
Technical field
The present invention relates to ad system technical fields, more particularly to a kind of information recommendation method and device.
Background technique
In ad system, system can usually obtain the demand intention of user by the analysis to user's routine behavior, from And the information for recommending some users to need to user.
In practical applications, ad system can obtain the information of user's needs according to the search behavior of user, wherein The search behavior of user includes: that user inputs search term and user in other search in the search function bar that ad system provides Search term etc. is inputted in engine, ad system generates recommendation information according to above-mentioned user behavior analysis result, and in ad system User is showed on screen.For example, user goes for the information about " school ", inputs and search in search box or search engine Rope word " school ", system can be analyzed according to the above-mentioned search behavior of user show that " school " is the information that user needs, so as to To recommend the relevant information about " school " to user.
The website information accessed by search behavior for user, since there is no search behavior, ad system without Method knows the intention of user, to can not recommend accurate information for user, so that waste system resource.
Summary of the invention
In view of the above problems, it proposes on the present invention overcomes the above problem or at least be partially solved in order to provide one kind State the information recommendation method and device of problem.
According to one aspect of the present invention, a kind of information recommendation method is provided, comprising:
According to the search behavior historical record of each user, the weight of each website information under each search term is counted；
According to the browsing behavior historical record of each user, the click of each website information under each user identifier is counted Number；
According to each website information under the number of clicks and each search term of each website information under each user identifier Weight, calculate the weight of each search term of each website information under each user identifier；
When client where a user identifier newly accesses a website information, by the recommendation request obtain user identifier and It is forward to extract sequence for website information, and the weight of each search term according to the website information under the user identifier At least one search term simultaneously generates recommendation information based on described search word；
The recommendation information to client is returned to be shown.
Preferably, the search behavior historical record according to each user, counts each network address under each search term The step of weight of information, comprising:
For the website information under a search term, by the number of clicks of the website information divided by institute under described search word There is the sum of the number of clicks of website information, obtains the first result；
First result is subjected to smoothing computation, obtains weight of the website information under described search word.
Preferably, described that first result is subjected to smoothing computation, the website information is obtained under described search word Weight the step of, comprising:
The sum of number of clicks and non-zero normal number to all website informations under described search word take logarithm, are smoothly joined Number；
First result is multiplied with smoothing parameter, obtains weight of the website information under described search word.
Preferably, when the client where a user identifier newly accesses a website information, under the user identifier The weight of each search term of the website information is extracted and sorts at least one forward search term and be based on the life of described search word The step of at recommendation information, comprising:
Receive the recommendation request that client is sent；The recommendation request is to access a net in client where a user identifier It is sent when the information of location to Advertisement Server；
User identifier and website information are obtained by the recommendation request, and believed according to the network address under the user identifier The weight of each search term of breath is extracted and sorts at least one forward search term and generate recommendation based on described search word Breath.
Preferably, described the step of recommendation information is generated based on described search word, comprising:
The advertisement link of described search word is corresponded to from advertisement library lookup；
Recommendation information is generated based on the advertisement link and search term.
Preferably, after returning to the step of recommendation information to client is shown, further includes:
Receive the access request triggered in the client to the clicking operation of a recommendation information by user；
According to the access request, corresponding advertisement page is returned to client.
It is preferably, described to return to the step of recommendation information is shown to client, comprising:
The recommendation information is returned to browser client, is shown with being generated in pop-up in browser client.
Preferably, the search behavior historical record according to each user, counts each network address under each search term The step of weight of information, comprising:
To the website information recorded under each search term in search history behavior record, according to content corresponding to website information Theme merger；
Count the weight of the website information under each search term after merger.
Preferably, the browsing behavior historical record according to each user, counts each net under each user identifier The step of number of clicks of location information, comprising:
To the website information in the browsing behavior historical record of each user, according to the theme of content corresponding to website information Merger；
Count the number of clicks of the website information under each user identifier after merger.
According to another aspect of the invention, a kind of information recommending apparatus is provided, comprising:
Website information weight statistical module counts each search for the search behavior historical record according to each user The weight of each website information under word；
Website information number of clicks statistical module, for the browsing behavior historical record according to each user, statistics is each The number of clicks of each website information under user identifier；
Search term weight calculation module, for according to the number of clicks of each website information under each user identifier and each The weight of each website information under search term calculates the power of each search term of each website information under each user identifier Weight；
Recommendation information generation module, for when client where a user identifier newly accesses a website information, according to institute The weight of each search term of the website information under user identifier is stated, at least one forward search term of sequence is extracted and is based on Described search word generates recommendation information；
Recommendation information display module is shown for returning to the recommendation information to client.
Preferably, the website information weight statistical module, comprising:
First result computational submodule, the website information for being directed under a search term, by the point of the website information Number is hit divided by the sum of the number of clicks of all website informations under described search word, obtains the first result；
The smooth submodule of first result obtains the website information and exists for first result to be carried out smoothing computation Weight under described search word.
Preferably, the smooth submodule of the first result, comprising:
Smoothing parameter acquisition submodule, for all website informations under described search word number of clicks and non-zero it is normal The sum of number takes logarithm, obtains smoothing parameter；
The first computational submodule of website information weight obtains described for first result to be multiplied with smoothing parameter Weight of the website information under described search word.
Preferably, the recommendation information generation module, comprising:
Recommendation request receiving submodule, for receiving the recommendation request of client transmission；The recommendation request is to use one Client where the mark of family is sent when accessing a website information to Advertisement Server；
Recommendation information first generates submodule, for obtaining user identifier and website information, and root by the recommendation request According to the weight of each search term of the website information under the user identifier, at least one the forward search term that sorts is extracted And recommendation information is generated based on described search word.
Preferably, the recommendation information first generates submodule, comprising:
Advertisement link searches submodule, for corresponding to the advertisement link of described search word from advertisement library lookup；
Recommendation information second generates submodule, for generating recommendation information based on the advertisement link and search term.
Preferably, after recommendation information display module, further includes:
Clicking operation receiving module in the client touches the clicking operation of a recommendation information by user for receiving The access request of hair；
Page jump module, for returning to corresponding advertisement page to client according to the access request.
Preferably, the recommendation information display module, comprising:
Pop-up shows submodule, for returning to the recommendation information to browser client, in browser client It generates and is shown in pop-up.
Preferably, the website information weight statistical module, comprising:
Website information the first merger submodule, for the network address recorded under each search term in search history behavior record Information, according to the theme merger of content corresponding to website information；
Website information weight merger submodule, for counting the weight of the website information under each search term after merger.
Preferably, the website information number of clicks statistical module, comprising:
Website information the second merger submodule, for the website information in the browsing behavior historical record to each user, According to the theme merger of content corresponding to website information；
Number of clicks merger submodule, for counting the number of clicks of the website information under each user identifier after merger.
It is off line can to count search term from search behavior historical record for information recommendation method and device according to the present invention The weight of location information, from the number of clicks of the website information under counting user mark in browsing behavior historical record.So it is directed to One user identifier can be according to each website information of each aforementioned obtained search term for each website information of its access Weight and the user identifier are associated with the access weight of each website information with website information, calculate each under the user identifier The weight of search term.So, when user is taken notice of with the user identifier and accesses some network address in client, ad system then can be with One or more forward search term of weight sequencing is extracted, and is thus solved according to the search term to user's recommendation information It is not the network address accessed by search behavior for user, since there is no search behavior, ad system can not know user's It is intended to, to achieve the case where can directly accessing user network address the problem of accurate information can not being recommended for user, look into Intention relevant to the network address is looked for, so as to recommend accurate information, reduces the beneficial effect of the waste of system resource.
The above description is only an overview of the technical scheme of the present invention, in order to better understand the technical means of the present invention, And it can be implemented in accordance with the contents of the specification, and in order to allow above and other objects of the present invention, feature and advantage can It is clearer and more comprehensible, the followings are specific embodiments of the present invention.
Detailed description of the invention
By reading the following detailed description of the preferred embodiment, various other advantages and benefits are common for this field Technical staff will become clear.The drawings are only for the purpose of illustrating a preferred embodiment, and is not considered as to the present invention Limitation.And throughout the drawings, the same reference numbers will be used to refer to the same parts.In the accompanying drawings:
Fig. 1 shows the step flow chart of information recommendation method embodiment one of the invention；
Fig. 2 shows the step flow charts of information recommendation method embodiment two of the invention；
Fig. 3 shows the step flow chart of information recommendation method embodiment three of the invention；
Fig. 4 shows the structural block diagram of information recommendation system example IV of the invention；
Fig. 5 shows the structural block diagram of information recommending apparatus embodiment five of the invention；
Fig. 6 shows the structural block diagram of information recommending apparatus embodiment six of the invention.
Specific embodiment
Exemplary embodiments of the present disclosure are described in more detail below with reference to accompanying drawings.Although showing the disclosure in attached drawing Exemplary embodiment, it being understood, however, that may be realized in various forms the disclosure without should be by embodiments set forth here It is limited.On the contrary, these embodiments are provided to facilitate a more thoroughly understanding of the present invention, and can be by the scope of the present disclosure It is fully disclosed to those skilled in the art.
Core of the invention thought is, from the weight for counting website information under search term in search behavior historical record, From the number of clicks of the website information under counting user mark in browsing behavior historical record.It is so directed to a user identifier, It, can be according to the weight and the use of each website information of each aforementioned obtained search term for each website information of its access Family identifies the access weight to each website information, is associated with website information, calculates the weight of each search term under the user identifier. So, take notice of that ad system can then extract weight sequencing when accessing some network address in client with the user identifier in user One or more forward search term, and according to the search term to user's recommendation information, thus solve is not for user The network address accessed by search behavior, since there is no search behavior, ad system can not know the intention of user, thus can not The problem of recommending accurate information for user achieves the case where can directly accessing user network address, searches and the network address phase The intention of pass reduces the beneficial effect of the waste of system resource so as to recommend accurate information.
Embodiment one
Referring to Fig.1, a kind of step flow chart of information recommendation method embodiment one according to the present invention is shown, specifically may be used To include the following steps:
Step 110, the search behavior historical record according to each user, count each website information under each search term Weight；
The system architecture of the embodiment of the present invention may include: client, ad system, third-party server, wherein client End can be browser or other can provide the software of browsing website information；Ad system is according to user to the clear of website information It lookes at behavior historical record and search behavior historical record calculates the weight of each website information under each search term and according to net The biggish search term of weight generates recommendation information under the information of location；Third-party server receives client and asks to the access of website information It asks, and returning response information after respective handling is done to access request.
It is appreciated that the system architecture of the embodiment of the present invention can also include search engine server, ad system can be with Search behavior historical record is obtained from search engine server.
In practical applications, user can be obtained by the function of search or other various search engines of ad system offer Information relevant to search term, therefore, user is every in a search engine once to be searched for, and search engine will record user's This time search behavior, after user carries out search behavior for several times, which can generate the search behavior historical record of user, So the server of ad system can obtain the search behavior historical record from the search engine.
Wherein, every record may include following information in search behavior historical record: user input search term and The website information that user clicks under search term.Wherein, website information can for URL (Uniform Resource Locator, Uniform resource locator) form, URL includes: resource type, host name, the resource file name for storing resource, wherein resources-type Type can refer to the domain name system host name or IP address of the server of storage resource, money for https agreement or http etc., host name Source filename refers to resource file title.
It is appreciated that in fact, being the history for clicking all users using the search term for a search term Record merges, and then counts the weight of each website information under the search term.For example, user A clicks net by search term W1 The number of location information N1, N2, N3 are respectively C1, C2, C3, and user B clicks the number difference of website information N3 by search term W1 For C4, user C is respectively C5, C6 by the number that search term W1 clicks website information N1, N2.Then, all user's uses are searched The historical record that rope word W1 clicks each website information obtains after merging: the number of clicks of website information N1 is use under search term W1 Family A passes through the number C5 of search term W1 click website information N1 by number C1, the user C that search term W1 clicks website information N1 The sum of, under search term W1 the number of clicks of website information N2 be user A by search term W1 click website information N2 number C2, User C clicks the sum of the number C6 of website information N2 by search term W1, and the number of clicks of website information N3 is under search term W1 User A passes through the number of search term W1 click website information N3 by number C3, the user B that search term W1 clicks website information N3 The sum of C4.
Preferably, step 110 may include sub-step 111 to 112:
Sub-step 111, for the website information under a search term, by the number of clicks of the website information divided by described The sum of the number of clicks of all website informations, obtains the first result under search term；
It can specifically be calculated using following formula:
Wherein, F (Qn,In,i) it is for n-th of search term Q in m search termnUnder i-th of website information In,iFirst As a result, certainly, the I under different search termsn,iCan be different, n is the integer for being less than or equal to m more than or equal to 1, knIt indicates n-th Search term QnUnder website information sum, it is whole more than or equal to 1 that the website information sum under different search terms can be different Number, C (Qn,In,i) indicate n-th of search term QnUnder i-th of website information In,iNumber of clicks.
As can be seen that first the result is that counted according to same user's difference website information, to n-th of search term Qn, the QnUnder i-th of website information In,iNumber of clicks C (Qn,In,i) bigger, the first result F (Qn,In,i) bigger, conversely, First result F (Qn,In,i) bigger, show n-th of search term QnUnder i-th of website information In,iNumber of clicks C (Qn,In,i) It is bigger.
First result is carried out smoothing computation by sub-step 112, obtains the website information under described search word Weight.
The number of clicks of website information the sum of of the first result that sub-step 111 obtains under two kinds of search terms differs greatly When will appear following situation, for example, the number of clicks of website information N1, N2 is 1 time under search term B1, then according to formula (1) The first result being calculated is 0.5, and the number of clicks of website information N1, N2 is respectively 10,900 times under search term B2, then It is respectively 0.1 and 0.9 according to the first result that formula (1) is calculated, at this point, first of the website information N1 under search term B1 As a result 0.5 is bigger than the first result 0.1 of the website information N1 under search term B2, but the website information N1 under actually search term B1 Number of clicks 1 but be far smaller than search term B2 under website information N1 number of clicks, to will lead to access search rate First result of very low network address is very big, but its contribution very little to search term, but its influence to weight is big, so the One result directly describes weight inaccuracy of the website information under search term, to solve the above-mentioned problems, needs by smooth Calculating is adjusted the first result, so that calculated result is more reasonable.
Power of the website information under described search word can be calculated according to the first result that sub-step 111 is calculated Weight.
Preferably, sub-step 112 may include sub-step 112A to 112B:
Sub-step 112A, the sum of the number of clicks of all website informations under described search word and non-zero normal number are taken pair Number, obtains smoothing parameter；
It can specifically be calculated using following formula:
Wherein, P (Qn,In,i) it is n-th of search term QnI-th of website information In,iSmoothing parameter, variable QnAnd In,i Meaning referring to described in formula (1), truth of a matter a can take any number for being more than or equal to 1 including 10 and e, and smoothing factor b can be with Take any number more than or equal to 1.
Sub-step 112B, first result is multiplied with smoothing parameter, obtains the website information in described search word Under weight.
It can specifically be calculated using following formula:
Wherein, V (Qn,Ii) it is n-th of search term QnI-th of website information In,iWeight, the meaning of other each variables Referring to described in formula (1) and formula (2).
Step 120, the browsing behavior historical record according to each user count each network address letter under each user identifier The number of clicks of breath；
Wherein, user identifier can be with the identity of unique identification active user, it will be understood that in the embodiment of the present invention, the visitor By taking browser client as an example, user identifier can be the user account in the Accounting system of browser, the browser at family end The Accounting system of client and is docked independently of the Accounting system of third-party server with ad system.Certainly, ad system It can be docked with the Accounting system of third-party server, directly use user of the user in the Accounting system of third-party server Account is user identifier.It is, of course, also possible to the use of the physical identity address of user terminal where client be directly user identifier. Certainly, the present invention to the concrete form of user identifier with no restrictions.
In practical applications, user can browse the website information of needs by various modes, comprising: defeated on a web browser The URL for entering website information clicks URL from the webpage of current presentation to jump directly to target website information etc., and user is clear Website information of looking at can also have other modes, and the present invention browses the mode of website information with no restrictions to user.So as to obtain Each user browsing historical behavior record, record in browsing historical behavior record to be that user is not hit by Searching point right The browsing behavior of network address.
In embodiments of the present invention, when user browses website information, it can be recorded by client and browses history row Then to there is client that the browsing historical behavior is sent to the server of advertisement.It can certainly be remembered by third-party server Family this time browsing behavior is employed, is then sent to the server of advertisement.The browsing historical behavior includes: user identifier, net Location information URL etc..
The browsing behavior historical record for the user A that client or server are generated by user identifier be (Ua, N1), (Ua, N2), (Ua, N3), (Ua, N3), (Ua, N4), (Ua, N2), (Ua, N3), (Ua, N4) }, the browsing behavior history of user B It is recorded as { (Ub, N1), (Ub, N3), (Ub, N2), (Ub, N1), (Ub, N4), (Ub, N1), (Ub, N3) }, wherein then user A Under the number of clicks of N1 website information be 1, the number of clicks of the N2 website information under user A is 2, the N3 network address under user A The number of clicks of information is 3, and the number of clicks of the N4 website information under user A is 2, the click of the N1 website information under user B Number is 3, and the number of clicks of the N2 website information under user B is 1, and the number of clicks of the N3 website information under user B is 2, is used The number of clicks of N4 website information under the B of family is 1.It is appreciated that if when a user does not access a website information, the network address The access times of information are 0.
Step 130, according to each under the number of clicks and each search term of each website information under each user identifier The weight of website information calculates the weight of each search term of each website information under each user identifier；
For each website information under all search terms described in step 110, calculated under each search term one by one Each website information can specifically calculate the weight of different user using following formula:
V(Ut,Qn,In,i)=C (Ut,In,i)*V(Qn,In,i) (4)
Wherein, V (Ut,Qn,In,i) it is t-th of user UtUnder n-th of search term i-th of website information In,iWeight, V(Qn,In,i) it is n-th of search term Q that formula (3) are calculatednLower i-th of website information In,iWeight, C (Ut,In,i) be T-th of user UtTo n-th of search term QnI-th of website information In,iNumber of clicks, ad system obtains from step 120 Website information mark and website information I are obtained in each website information number of clicks result under each user identifiern,iNetwork address It identifies and identical records corresponding number of clicks as n-th of search term QnLower i-th of website information In,iNumber of clicks.
Step 140, when client where a user identifier newly accesses a website information, according under the user identifier The weight of each search term of the website information is extracted and sorts at least one forward search term and be based on the generation of described search word Recommendation information；
Wherein, client can be to be mounted on terminal device that the software of website information access function, packet can be provided Include: browser or other software of website information access function can be provided, terminal device includes desktop computer, notebook electricity Brain, mobile phone etc., the present invention to the form of client and terminal device with no restrictions.
Certainly, client of the invention is not when clicking a website information by search results pages, to third-party server The request of access website information is sent, includes user identifier in the request.Third-party server marks the website information and user Knowledge is sent to Advertisement Server.The sequence that Advertisement Server then searches the website information under the user identifier of record is forward Then one or more search term generates recommendation information according to these search terms and then returns to client, client can open up Show the recommendation information.
Advertisement Server of the invention obtains that above-mentioned user identifier is corresponding all to search from the search term that step 130 obtains Rope word, and search term is arranged according to weight descending, the forward search term that sorts is extracted, described search word and search term is corresponding Link as a recommendation information.Wherein, since the biggish search term of weight is related compared with the browsing behavior of user, thus User demand is more met by the recommendation information that the biggish search term of weight generates.
Step 150, the return recommendation information to client are shown.
Server can return to the recommendation information after generating recommendation information, and then client can be in designated position Show the recommendation information.
Wherein, designated position can be any region of client screen, but answer in order to not influence user is currently in use With the window of recommendation information being preferably placed in screen edges and corners.Show that the mode of recommendation information can specifically include: Search term and the corresponding advertisement chain link of search term are presented in system screen, if it exists multiple search terms, then according to search The weight of word is displayed on the screen according to sequence arrangement from big to small, at this point it is possible to the maximum value of recommendation information number is set, When the number of recommendation information is more than or equal to above-mentioned maximum value, the recommendation information number of displaying is maximum value.
Preferably, the non-browser client of the recommendation information is received, then pop-up can be generated in browser client, and Above-mentioned recommendation information is shown among the pop-up, wherein the size of pop-up can be closed according to the size of system screen The setting of reason, under normal conditions preferably no more than the 10 of screen, the position of pop-up can be the lower left corner, the bottom right of screen Angle or other do not interfere the position etc. of user's normal use system, can be appropriate when needing to show there are multiple recommendation informations The wicket of recommendation information is reduced, and wicket is shown according to lateral, longitudinal or other layouts on the screen, in reality In, when the number of search term is greater than 1, the number of corresponding recommendation information is also greater than 1, then according to the weight of search term by big The corresponding recommendation information of search term is shown to small sequence.It is appreciated that can also have other modes to show recommendation information, originally Invention is to showing the mode of recommendation information, position, size, shape, layout with no restrictions.
When occurring the window of recommendation information on screen, user can according to need selection and check or close.In client Side,
The access request that step 160, reception in the client trigger the clicking operation of a recommendation information by user；
Step 170, according to the access request, return to corresponding advertisement page to client.
If user needs to check certain recommendation information, a recommendation information shown in pop-up can be clicked.User's point When hitting recommendation information access request would generally be sent to Advertisement Server.Due to the embodiment of the present invention to every recommendation information rear Platform has hung link, then can call browser kernel that access request occurs to Advertisement Server when user clicks recommendation information, because This access request is actually a web-page requests.After Advertisement Server receives the access request, then according in the request Link return to corresponding advertisement page to the client, which then shows the advertisement page.
The embodiment of the present invention can be from the weight for counting website information under search term in search behavior historical record, from browsing The number of clicks of website information in behavior historical record under counting user mark.It is so directed to a user identifier, for it Each website information of access, can be according to the weight and the user identifier of each website information of each aforementioned obtained search term It to the access weight of each website information, is associated with website information, calculates the weight of each search term under the user identifier.So, When user is taken notice of with the user identifier and accesses some network address in client, it is forward that Advertisement Server can then extract weight sequencing One or more search term thus solve and user do not passed through and according to the search term to user's recommendation information The network address of search behavior access, since there is no search behavior, Advertisement Server can not know the intention of user, to can not be User recommends the problem of accurate information, achieves the case where can directly accessing user network address, searches related to the network address Intention reduce the beneficial effect of the waste of system resource so as to recommend accurate information.
Embodiment two
Referring to Fig. 2, a kind of information recommendation method embodiment two according to the present invention is shown, specific steps may include:
Step 210, to the website information recorded under each search term in search history behavior record, according to website information institute The theme merger of corresponding content；
Under normal conditions, the corresponding different website informations of the same search term are to be mutually related, the content of website information Belong to same subject, in this case, in order to reduce calculation amount, can first to search behavior historical record according to theme into Row merger, so that the quantity for participating in the website information calculated is reduced.
In the multiple website informations recorded under one search term, the content of certain a few website information is correlation or identical, then It can be a website information group by these network address merger, multiple network address in group are equal to the number of clicks of the website information group and are believed The sum of number of clicks of breath.For example, the corresponding number of clicks point of website information A1, A2, A3, A4, A5 for being recorded under a search term It Wei C1, C2, C3, C4, C5, wherein website information A1 is the Teaching Advantages about school S in chemical aspect, and A3 is about Advantage of the school S in terms of scientific research, A4 are the advantages about school S in terms of educational alternative, these three website information corresponding contents Belong to same subject, be all the advantage about school S, is then website information group D, a D institute by website information A1, A3 and A4 merger Corresponding number of clicks is the sum of C1, C3, C4 C, and the result obtained after merger is D, A2, A5, and corresponding number of clicks is respectively C、C2、C5。
Obtain website information content theme method may include: Statistics-Based Method, Knowledge based engineering method with And the method combined based on the two, wherein Statistics-Based Method utilizes the frequency of characteristic item, the position of characteristic item, characteristic item Contribution information etc. obtain theme, Knowledge based engineering method carries out grammer language using knowledge base such as script, machine readable dictionary Justice analysis improves overall performance by combining two methods to obtain theme, based on the method that the two combines.The present invention couple Obtain the subject methods of website information content with no restrictions.
The weight of website information under step 220, each search term of statistics after merger.
Above-mentioned search term is calculated according still further to the sub-step 111,112 of embodiment one kind on the basis of website information after merger The weight of lower website information.
Step 230, to the website information in the browsing behavior historical record of each user, in corresponding to website information The theme merger of appearance；
Due to being carried out to website information according to the theme of corresponding content in the weight of website information under calculating search term Merger, similarly, in the case where calculating user identifier, the number of clicks of website information is also required to the master to website information according to corresponding content Topic carries out same merger, so that website information and search history behavior record in the browsing behavior historical record of user Website information under middle search term corresponds to each other.
The number of clicks of website information under step 240, each user identifier of statistics after merger.
After multiple website informations in system under a user identifier are according to the theme merger of corresponding content, total number of clicks etc. The sum of the number of clicks of multiple website informations in organizing.
Step 250, according to each under the number of clicks and each search term of each website information under each user identifier The weight of website information calculates the weight of each search term of each website information under each user identifier；
Step 260, when client where a user identifier newly accesses a website information, according under the user identifier The weight of each search term of the website information is extracted and sorts at least one forward search term and be based on the generation of described search word Recommendation information；
Step 270, the return recommendation information to client are shown.
The embodiment of the present invention can be from the weight for counting website information under search term in search behavior historical record, from browsing The number of clicks of website information in behavior historical record under counting user mark.It is so directed to a user identifier, for it Each website information of access, can be according to the weight and the user identifier of each website information of each aforementioned obtained search term It to the access weight of each website information, is associated with website information, calculates the weight of each search term under the user identifier.So, When user is taken notice of with the user identifier and accesses some network address in client, it is forward that Advertisement Server can then extract weight sequencing One or more search term thus solve and user do not passed through and according to the search term to user's recommendation information The network address of search behavior access, since there is no search behavior, Advertisement Server can not know the intention of user, to can not be User recommends the problem of accurate information, achieves the case where can directly accessing user network address, searches related to the network address Intention reduce the beneficial effect of the waste of system resource so as to recommend accurate information.
Also, the embodiment of the present invention is to the website information that records under each search term in search history behavior record and each Website information in the browsing behavior historical record of user, according to the theme merger of content corresponding to website information, to reduce The point of website information under the weight and each user identifier of website information under each search term of statistics after merger after merger The calculation amount for hitting number finally improves the efficiency of information recommendation.
Embodiment three
Referring to Fig. 3, a kind of information recommendation method embodiment three according to the present invention is shown, specific steps may include:
Step 310, the search behavior historical record according to each user, count each website information under each search term Weight；
Step 320, the browsing behavior historical record according to each user count each network address letter under each user identifier The number of clicks of breath；
Step 330, according to each under the number of clicks and each search term of each website information under each user identifier The weight of website information calculates the weight of each search term of each website information under each user identifier；
Step 340 receives the recommendation request that client is sent；The recommendation request is in client where a user identifier It is sent when accessing a website information to Advertisement Server；
When clicking website information in user in the client non-search result page, client can be to third Square server sends access request, and third-party server can be various application servers.Meanwhile client is to Advertisement Server Send the recommendation request of recommendation information.Or third-party server sends recommendation request, the recommendation request packet to Advertisement Server The physical address information of client, such as IP address are included, Advertisement Server is facilitated to connect client.
Step 350 obtains user identifier and website information by the recommendation request, and according to the institute under the user identifier The weight of each search term of website information is stated, is extracted and is sorted at least one forward search term and be based on the generation of described search word Recommendation information；
Client-server initiates the recommendation request of recommendation information to Advertisement Server, wherein the recommendation of recommendation information is asked User identifier and website information URL etc. are carried in asking.Then server can from the recommendation request resolve user identity And website information is extracted sequence and is leaned on then according to the weight of each search term of the website information under the user identifier At least one preceding search term simultaneously generates recommendation information based on described search word.
If sending recommendation request by third-party server, wherein can also include the IP address of client.Preferably, it walks The step of generating recommendation information based on described search word in rapid 350, may include sub-step 351 to 352:
Sub-step 351, the advertisement link that described search word is corresponded to from advertisement library lookup；
Sub-step 352 generates recommendation information based on the advertisement link and search term.
Wherein, advertisement base is the database for storing all search terms and its corresponding advertisement link, can be existing each The database of seed type, the present invention to the type of database with no restrictions.In practical application, search term is established index first, Search term and index are compared one by one in retrieving, if search term is matched with certain index, then will currently be indexed corresponding Advertisement link of the advertisement link as search term, the link and search term are as one group of recommendation information.
Step 360, the return recommendation information to client are shown.
Recommendation request is directly transmitted to Advertisement Server if it is client, then client is directly established with Advertisement Server Link, Advertisement Server directly can return to the recommendation information according to linking for client.
It is sent if it is third-party server to Advertisement Server and recommends to ask, the Advertisement Server of the embodiment of the present invention then may be used Established the link with client according to the IP address of the client in recommendation request, the recommendation information is then returned into client.
The embodiment of the present invention can be from the weight for counting website information under search term in search behavior historical record, from browsing The number of clicks of website information in behavior historical record under counting user mark.It is so directed to a user identifier, for it Each website information of access, can be according to the weight and the user identifier of each website information of each aforementioned obtained search term It to the access weight of each website information, is associated with website information, calculates the weight of each search term under the user identifier.So, When user is taken notice of with the user identifier and accesses some network address in client, it is forward that Advertisement Server can then extract weight sequencing One or more search term thus solve and user do not passed through and according to the search term to user's recommendation information The network address of search behavior access, since there is no search behavior, Advertisement Server can not know the intention of user, to can not be User recommends the problem of accurate information, achieves the case where can directly accessing user network address, searches related to the network address Intention reduce the beneficial effect of the waste of system resource so as to recommend accurate information.
Also, the embodiment of the present invention is by combining search term and the corresponding advertisement link of search term as recommendation Breath, so that user can understand the theme of recommendation information according to the search term in recommendation information, clicking search term can directly be jumped Advertisement page is gone to, reduces the operating procedure that user opens recommendation information, is brought conveniently for user.
For embodiment of the method, for simple description, therefore, it is stated as a series of action combinations, but this field Technical staff should be aware of, and embodiment of that present invention are not limited by the describe sequence of actions, because implementing according to the present invention Example, some steps may be performed in other sequences or simultaneously.Secondly, those skilled in the art should also know that, specification Described in embodiment belong to preferred embodiment, the actions involved are not necessarily necessary for embodiments of the present invention.
Example IV
Referring to Fig. 4, a kind of structural block diagram of information recommending apparatus example IV according to the present invention is shown, specifically can wrap Include following module:
Website information weight statistical module 410, for the search behavior historical record according to each user, statistics is each searched The weight of each website information under rope word；
Website information number of clicks statistical module 420, for the browsing behavior historical record according to each user, statistics is every The number of clicks of each website information under a user identifier；
Search term weight calculation module 430, for according to the number of clicks of each website information under each user identifier and The weight of each website information under each search term calculates each search term of each website information under each user identifier Weight；
Recommendation information generation module 440, for when client where a user identifier newly accesses a website information, according to The weight of each search term of the website information under the user identifier extracts sort at least one forward search term and base Recommendation information is generated in described search word；
Recommendation information display module 450 is shown for returning to the recommendation information to client.
Preferably, the website information weight statistical module, comprising:
First result computational submodule, the website information for being directed under a search term, by the point of the website information Number is hit divided by the sum of the number of clicks of all website informations under described search word, obtains the first result；
The smooth submodule of first result obtains the website information and exists for first result to be carried out smoothing computation Weight under described search word.
Preferably, the smooth submodule of the first result, comprising:
Smoothing parameter acquisition submodule, for all website informations under described search word number of clicks and non-zero it is normal The sum of number takes logarithm, obtains smoothing parameter；
The first computational submodule of website information weight obtains described for first result to be multiplied with smoothing parameter Weight of the website information under described search word.
Preferably, the recommendation information generation module, comprising:
Recommendation request receiving submodule, for receiving the recommendation request of client transmission；The recommendation request is to use one Client where the mark of family is sent when accessing a website information to Advertisement Server；
Recommendation information first generates submodule, for obtaining user identifier and website information, and root by the recommendation request According to the weight of each search term of the website information under the user identifier, at least one the forward search term that sorts is extracted And recommendation information is generated based on described search word.
Preferably, the recommendation information first generates submodule, comprising:
Advertisement link searches submodule, for corresponding to the advertisement link of described search word from advertisement library lookup；
Recommendation information second generates submodule, for generating recommendation information based on the advertisement link and search term.
Preferably, after recommendation information display module, further includes:
Clicking operation receiving module in the client touches the clicking operation of a recommendation information by user for receiving The access request of hair；
Page jump module, for returning to corresponding advertisement page to client according to the access request.
Preferably, the recommendation information display module, comprising:
Pop-up shows submodule, for returning to the recommendation information to browser client, in browser client It generates and is shown in pop-up.
Preferably, the website information weight statistical module, comprising:
Website information the first merger submodule, for the network address recorded under each search term in search history behavior record Information, according to the theme merger of content corresponding to website information；
Website information weight merger submodule, for counting the weight of the website information under each search term after merger.
Preferably, the website information number of clicks statistical module, comprising:
Website information the second merger submodule, for the website information in the browsing behavior historical record to each user, According to the theme merger of content corresponding to website information；
Number of clicks merger submodule, for counting the number of clicks of the website information under each user identifier after merger.
The embodiment of the present invention can be from the weight for counting website information under search term in search behavior historical record, from browsing The number of clicks of website information in behavior historical record under counting user mark.It is so directed to a user identifier, for it Each website information of access, can be according to the weight and the user identifier of each website information of each aforementioned obtained search term It to the access weight of each website information, is associated with website information, calculates the weight of each search term under the user identifier.So, When user is taken notice of with the user identifier and accesses some network address in client, it is forward that Advertisement Server can then extract weight sequencing One or more search term thus solve and user do not passed through and according to the search term to user's recommendation information The network address of search behavior access, since there is no search behavior, Advertisement Server can not know the intention of user, to can not be User recommends the problem of accurate information, achieves the case where can directly accessing user network address, searches related to the network address Intention reduce the beneficial effect of the waste of system resource so as to recommend accurate information.
Embodiment five
Referring to Fig. 5, a kind of structural block diagram of information recommending apparatus example IV according to the present invention is shown, specifically can wrap Include following module:
Website information weight statistical module 510, for the search behavior historical record according to each user, statistics is each searched The weight of each website information under rope word；It specifically includes:
The first merger of website information submodule 511, for recording under each search term in search history behavior record Website information, according to the theme merger of content corresponding to website information；
Website information weight merger submodule 512, for counting the weight of the website information under each search term after merger.
Website information number of clicks statistical module 520, for the browsing behavior historical record according to each user, statistics is every The number of clicks of each website information under a user identifier；It specifically includes:
The second merger of website information submodule 521, for the network address letter in the browsing behavior historical record to each user Breath, according to the theme merger of content corresponding to website information；
Number of clicks merger submodule 522, for counting the click time of the website information under each user identifier after merger Number.
Search term weight calculation module 530, for according to the number of clicks of each website information under each user identifier and The weight of each website information under each search term calculates each search term of each website information under each user identifier Weight；
Recommendation information generation module 540, for when client where a user identifier newly accesses a website information, according to The weight of each search term of the website information under the user identifier extracts sort at least one forward search term and base Recommendation information is generated in described search word；
Recommendation information display module 550 is shown for returning to the recommendation information to client.
The embodiment of the present invention can be from the weight for counting website information under search term in search behavior historical record, from browsing The number of clicks of website information in behavior historical record under counting user mark.It is so directed to a user identifier, for it Each website information of access, can be according to the weight and the user identifier of each website information of each aforementioned obtained search term It to the access weight of each website information, is associated with website information, calculates the weight of each search term under the user identifier.So, When user is taken notice of with the user identifier and accesses some network address in client, it is forward that Advertisement Server can then extract weight sequencing One or more search term thus solve and user do not passed through and according to the search term to user's recommendation information The network address of search behavior access, since there is no search behavior, Advertisement Server can not know the intention of user, to can not be User recommends the problem of accurate information, achieves the case where can directly accessing user network address, searches related to the network address Intention reduce the beneficial effect of the waste of system resource so as to recommend accurate information.
Also, the embodiment of the present invention is to the website information that records under each search term in search history behavior record and each Website information in the browsing behavior historical record of user, according to the theme merger of content corresponding to website information, to reduce The point of website information under the weight and each user identifier of website information under each search term of statistics after merger after merger The calculation amount for hitting number finally improves the efficiency of information recommendation.
Embodiment six
Referring to Fig. 6, a kind of structural block diagram of information recommending apparatus example IV according to the present invention is shown, specifically can wrap Include following module:
Website information weight statistical module 610, for the search behavior historical record according to each user, statistics is each searched The weight of each website information under rope word；
Website information number of clicks statistical module 620, for the browsing behavior historical record according to each user, statistics is every The number of clicks of each website information under a user identifier；
Search term weight calculation module 630, for according to the number of clicks of each website information under each user identifier and The weight of each website information under each search term calculates each search term of each website information under each user identifier Weight；
Recommendation information generation module 640, for when client where a user identifier newly accesses a website information, according to The weight of each search term of the website information under the user identifier extracts sort at least one forward search term and base Recommendation information is generated in described search word；It specifically includes:
Recommendation request receiving submodule 641, for receiving the recommendation request of client transmission；The recommendation request is one Client where user identifier is sent when accessing a website information to Advertisement Server；
Recommendation information first generates submodule 642, for obtaining user identifier and website information by the recommendation request, and According to the weight of each search term of the website information under the user identifier, at least one search for sorting forward is extracted Word simultaneously generates recommendation information based on described search word.
Recommendation information display module 650 is shown for returning to the recommendation information to client.
The embodiment of the present invention can be from the weight for counting website information under search term in search behavior historical record, from browsing The number of clicks of website information in behavior historical record under counting user mark.It is so directed to a user identifier, for it Each website information of access, can be according to the weight and the user identifier of each website information of each aforementioned obtained search term It to the access weight of each website information, is associated with website information, calculates the weight of each search term under the user identifier.So, When user is taken notice of with the user identifier and accesses some network address in client, it is forward that Advertisement Server can then extract weight sequencing One or more search term thus solve and user do not passed through and according to the search term to user's recommendation information The network address of search behavior access, since there is no search behavior, Advertisement Server can not know the intention of user, to can not be User recommends the problem of accurate information, achieves the case where can directly accessing user network address, searches related to the network address Intention reduce the beneficial effect of the waste of system resource so as to recommend accurate information.
And the embodiment of the present invention is by combining search term and the corresponding advertisement link of search term as recommendation Breath, so that user can understand the theme of recommendation information according to the search term in recommendation information, clicking search term can directly be jumped Advertisement page is gone to, reduces the operating procedure that user opens recommendation information, is brought conveniently for user.
Algorithm and display are not inherently related to any particular computer, virtual system, or other device provided herein. Various general-purpose systems can also be used together with teachings based herein.As described above, it constructs required by this kind of system Structure be obvious.In addition, the present invention is also not directed to any particular programming language.It should be understood that can use various Programming language realizes summary of the invention described herein, and the description done above to language-specific is to disclose this hair Bright preferred forms.
In the instructions provided here, numerous specific details are set forth.It is to be appreciated, however, that implementation of the invention Example can be practiced without these specific details.In some instances, well known method, structure is not been shown in detail And technology, so as not to obscure the understanding of this specification.
Similarly, it should be understood that in order to simplify the disclosure and help to understand one or more of the various inventive aspects, Above in the description of exemplary embodiment of the present invention, each feature of the invention is grouped together into single implementation sometimes In example, figure or descriptions thereof.However, the disclosed method should not be interpreted as reflecting the following intention: i.e. required to protect Shield the present invention claims features more more than feature expressly recited in each claim.More precisely, as following Claims reflect as, inventive aspect is all features less than single embodiment disclosed above.Therefore, Thus the claims for following specific embodiment are expressly incorporated in the specific embodiment, wherein each claim itself All as a separate embodiment of the present invention.
Those skilled in the art will understand that can be carried out adaptively to the module in the equipment in embodiment Change and they are arranged in one or more devices different from this embodiment.It can be the module or list in embodiment Member or component are combined into a module or unit or component, and furthermore they can be divided into multiple submodule or subelement or Sub-component.Other than such feature and/or at least some of process or unit exclude each other, it can use any Combination is to all features disclosed in this specification (including adjoint claim, abstract and attached drawing) and so disclosed All process or units of what method or apparatus are combined.Unless expressly stated otherwise, this specification is (including adjoint power Benefit require, abstract and attached drawing) disclosed in each feature can carry out generation with an alternative feature that provides the same, equivalent, or similar purpose It replaces.
In addition, it will be appreciated by those of skill in the art that although some embodiments described herein include other embodiments In included certain features rather than other feature, but the combination of the feature of different embodiments mean it is of the invention Within the scope of and form different embodiments.For example, in the following claims, embodiment claimed is appointed Meaning one of can in any combination mode come using.
Various component embodiments of the invention can be implemented in hardware, or to run on one or more processors Software module realize, or be implemented in a combination thereof.It will be understood by those of skill in the art that can be used in practice Microprocessor or digital signal processor (DSP) come realize some in information recommendation equipment according to an embodiment of the present invention or The some or all functions of person's whole component.The present invention is also implemented as one for executing method as described herein Point or whole device or device programs (for example, computer program and computer program product).Such this hair of realization Bright program can store on a computer-readable medium, or may be in the form of one or more signals.It is such Signal can be downloaded from an internet website to obtain, and is perhaps provided on the carrier signal or is provided in any other form.
It should be noted that the above-mentioned embodiments illustrate rather than limit the invention, and ability Field technique personnel can be designed alternative embodiment without departing from the scope of the appended claims.In the claims, Any reference symbol between parentheses should not be configured to limitations on claims.Word "comprising" does not exclude the presence of not Element or step listed in the claims.Word "a" or "an" located in front of the element does not exclude the presence of multiple such Element.The present invention can be by means of including the hardware of several different elements and being come by means of properly programmed computer real It is existing.In the unit claims listing several devices, several in these devices can be through the same hardware branch To embody.The use of word first, second, and third does not indicate any sequence.These words can be explained and be run after fame Claim.
The invention discloses A1, a kind of information recommendation method, comprising:
According to the search behavior historical record of each user, the weight of each website information under each search term is counted；
According to the browsing behavior historical record of each user, the click of each website information under each user identifier is counted Number；
According to each website information under the number of clicks and each search term of each website information under each user identifier Weight, calculate the weight of each search term of each website information under each user identifier；
When client where a user identifier newly accesses a website information, by the recommendation request obtain user identifier and It is forward to extract sequence for website information, and the weight of each search term according to the website information under the user identifier At least one search term simultaneously generates recommendation information based on described search word；
The recommendation information to client is returned to be shown.
A2, method as described in a1, the search behavior historical record according to each user, count under each search term Each website information weight the step of, comprising:
For the website information under a search term, by the number of clicks of the website information divided by institute under described search word There is the sum of the number of clicks of website information, obtains the first result；
First result is subjected to smoothing computation, obtains weight of the website information under described search word.
A3, as described in A2 method, it is described that first result is subjected to smoothing computation, the website information is obtained in institute The step of stating the weight under search term, comprising:
The sum of number of clicks and non-zero normal number to all website informations under described search word take logarithm, are smoothly joined Number；
First result is multiplied with smoothing parameter, obtains weight of the website information under described search word.
A4, the method as described in one of A1-A3, the client where a user identifier newly access network address letter When breath, the weight of each search term of the website information under the user identifier is extracted at least one for sorting forward and is searched Rope word is simultaneously based on the step of described search word generates recommendation information, comprising:
Receive the recommendation request that client is sent；The recommendation request is to access a net in client where a user identifier It is sent when the information of location to Advertisement Server；
User identifier and website information are obtained by the recommendation request, and believed according to the network address under the user identifier The weight of each search term of breath is extracted and sorts at least one forward search term and generate recommendation based on described search word Breath.
A5, the method as described in A4, described the step of recommendation information is generated based on described search word, comprising:
The advertisement link of described search word is corresponded to from advertisement library lookup；
Recommendation information is generated based on the advertisement link and search term.
A6, method as described in a1, after returning to the step of recommendation information to client is shown, further includes:
Receive the access request triggered in the client to the clicking operation of a recommendation information by user；
According to the access request, corresponding advertisement page is returned to client.
A7, method as described in a1, it is described to return to the step of recommendation information is shown to client, comprising:
The recommendation information is returned to browser client, is shown with being generated in pop-up in browser client.
A8, the method as described in 1, the search behavior historical record according to each user, count under each search term Each website information weight the step of, comprising:
To the website information recorded under each search term in search history behavior record, according to content corresponding to website information Theme merger；
Count the weight of the website information under each search term after merger.
A9, the method as described in A8, the browsing behavior historical record according to each user, count each user identifier Under each website information number of clicks the step of, comprising:
To the website information in the browsing behavior historical record of each user, according to the theme of content corresponding to website information Merger；
Count the number of clicks of the website information under each user identifier after merger.
The invention also discloses B10, a kind of information recommending apparatus, comprising:
Website information weight statistical module counts each search for the search behavior historical record according to each user The weight of each website information under word；
Website information number of clicks statistical module, for the browsing behavior historical record according to each user, statistics is each The number of clicks of each website information under user identifier；
Search term weight calculation module, for according to the number of clicks of each website information under each user identifier and each The weight of each website information under search term calculates the power of each search term of each website information under each user identifier Weight；
Recommendation information generation module, for when client where a user identifier newly accesses a website information, according to institute The weight of each search term of the website information under user identifier is stated, at least one forward search term of sequence is extracted and is based on Described search word generates recommendation information；
Recommendation information display module is shown for returning to the recommendation information to client.
B11, the device as described in B10, the website information weight statistical module, comprising:
First result computational submodule, the website information for being directed under a search term, by the point of the website information Number is hit divided by the sum of the number of clicks of all website informations under described search word, obtains the first result；
The smooth submodule of first result obtains the website information and exists for first result to be carried out smoothing computation Weight under described search word.
B12, device as described in b11, the smooth submodule of the first result, comprising:
Smoothing parameter acquisition submodule, for all website informations under described search word number of clicks and non-zero it is normal The sum of number takes logarithm, obtains smoothing parameter；
The first computational submodule of website information weight obtains described for first result to be multiplied with smoothing parameter Weight of the website information under described search word.
B13, the device as described in one of 10-12, the recommendation information generation module, comprising:
Recommendation request receiving submodule, for receiving the recommendation request of client transmission；The recommendation request is to use one Client where the mark of family is sent when accessing a website information to Advertisement Server；
Recommendation information first generates submodule, for obtaining user identifier and website information, and root by the recommendation request According to the weight of each search term of the website information under the user identifier, at least one the forward search term that sorts is extracted And recommendation information is generated based on described search word.
B14, the device as described in 13, the recommendation information first generate submodule, comprising:
Advertisement link searches submodule, for corresponding to the advertisement link of described search word from advertisement library lookup；
Recommendation information second generates submodule, for generating recommendation information based on the advertisement link and search term.
B15, the device as described in B10, after recommendation information display module, further includes:
Clicking operation receiving module in the client touches the clicking operation of a recommendation information by user for receiving The access request of hair；
Page jump module, for returning to corresponding advertisement page to client according to the access request.
B16, the device as described in B10, the recommendation information display module, comprising:
Pop-up shows submodule, for returning to the recommendation information to browser client, in browser client It generates and is shown in pop-up.
B17, the device as described in B10, the website information weight statistical module, comprising:
Website information the first merger submodule, for the network address recorded under each search term in search history behavior record Information, according to the theme merger of content corresponding to website information；
Website information weight merger submodule, for counting the weight of the website information under each search term after merger.
B18, the device as described in B17, the website information number of clicks statistical module, comprising:
Website information the second merger submodule, for the website information in the browsing behavior historical record to each user, According to the theme merger of content corresponding to website information；
Number of clicks merger submodule, for counting the number of clicks of the website information under each user identifier after merger.

Claims (14)

1. a kind of information recommendation method, comprising:
According to the search behavior historical record of each user, the weight of each website information under each search term is counted；
According to the browsing behavior historical record of each user, the click time of each website information under each user identifier is counted Number；
According to the power of each website information under the number of clicks and each search term of each website information under each user identifier Weight, calculates the weight of each search term of each website information under each user identifier；
When client where a user identifier newly accesses a website information, user identifier and network address are obtained by the recommendation request It is forward at least to extract sequence for information, and the weight of each search term according to the website information under the user identifier One search term simultaneously generates recommendation information based on described search word；
The recommendation information to client is returned to be shown；
The website information when client where a user identifier newly accesses a website information, under the user identifier Each search term weight, extract forward at least one search term of sorting and recommendation information simultaneously generated based on described search word Step, comprising:
Receive the recommendation request that client is sent；The recommendation request is to access network address letter in client where a user identifier It is sent when breath to Advertisement Server；
User identifier and website information are obtained by the recommendation request, and according to the website information under the user identifier The weight of each search term is extracted and sorts at least one forward search term and generate recommendation information based on described search word；
Described the step of recommendation information is generated based on described search word, comprising:
The advertisement link of described search word is corresponded to from advertisement library lookup；
Recommendation information is generated based on the advertisement link and search term.
2. the method according to claim 1, wherein the search behavior historical record according to each user, The step of counting the weight of each website information under each search term, comprising:
For the website information under a search term, by the number of clicks of the website information divided by all nets under described search word The sum of the number of clicks of location information, obtains the first result；
First result is subjected to smoothing computation, obtains weight of the website information under described search word.
3. according to the method described in claim 2, it is characterized in that, it is described by first result carry out smoothing computation, obtain The step of weight of the website information under described search word, comprising:
The sum of number of clicks and non-zero normal number to all website informations under described search word take logarithm, obtain smoothing parameter；
First result is multiplied with smoothing parameter, obtains weight of the website information under described search word.
4. the method according to claim 1, wherein returning to the step that the recommendation information to client is shown After rapid, further includes:
Receive the access request triggered in the client to the clicking operation of a recommendation information by user；
According to the access request, corresponding advertisement page is returned to client.
5. the method according to claim 1, wherein the recommendation information to the client that returns is shown The step of, comprising:
The recommendation information is returned to browser client, is shown with being generated in pop-up in browser client.
6. the method according to claim 1, wherein the search behavior historical record according to each user, The step of counting the weight of each website information under each search term, comprising:
To the website information recorded under each search term in search history behavior record, according to the master of content corresponding to website information Inscribe merger；
Count the weight of the website information under each search term after merger.
7. according to the method described in claim 6, it is characterized in that, the browsing behavior historical record according to each user, The step of counting the number of clicks of each website information under each user identifier, comprising:
To the website information in the browsing behavior historical record of each user, return according to the theme of content corresponding to website information And；
Count the number of clicks of the website information under each user identifier after merger.
8. a kind of information recommending apparatus, comprising:
Website information weight statistical module counts under each search term for the search behavior historical record according to each user Each website information weight；
Website information number of clicks statistical module counts each user for the browsing behavior historical record according to each user The number of clicks of each website information under mark；
Search term weight calculation module, for the number of clicks and each search according to each website information under each user identifier The weight of each website information under word calculates the weight of each search term of each website information under each user identifier；
Recommendation information generation module, for when client where a user identifier newly accesses a website information, according to the use The weight of each search term of the website information under the mark of family extracts at least one forward search term of sequence and based on described Search term generates recommendation information；
Recommendation information display module is shown for returning to the recommendation information to client；
The recommendation request is to send when client where a user identifier accesses a website information to Advertisement Server；
Recommendation information first generates submodule, for obtaining user identifier and website information by the recommendation request, and according to institute The weight of each search term of the website information under user identifier is stated, sort at least one forward search term and base are extracted Recommendation information is generated in described search word；
The recommendation information first generates submodule, comprising:
Advertisement link searches submodule, for corresponding to the advertisement link of described search word from advertisement library lookup；
Recommendation information second generates submodule, for generating recommendation information based on the advertisement link and search term.
9. device according to claim 8, which is characterized in that the website information weight statistical module, comprising:
First result computational submodule, the website information for being directed under a search term, by the click of the website information time Number obtains the first result divided by the sum of the number of clicks of all website informations under described search word；
The smooth submodule of first result obtains the website information described for first result to be carried out smoothing computation Weight under search term.
10. device according to claim 9, which is characterized in that the smooth submodule of the first result, comprising:
Smoothing parameter acquisition submodule, for all website informations under described search word number of clicks and non-zero normal number it With take logarithm, obtain smoothing parameter；
The first computational submodule of website information weight obtains the network address for first result to be multiplied with smoothing parameter Weight of the information under described search word.
11. device according to claim 8, which is characterized in that after recommendation information display module, further includes:
Clicking operation receiving module by user in the client triggers the clicking operation of a recommendation information for receiving Access request；
Page jump module, for returning to corresponding advertisement page to client according to the access request.
12. device according to claim 8, which is characterized in that the recommendation information display module, comprising:
Pop-up shows submodule, for returning to the recommendation information to browser client, to generate in browser client It is shown in pop-up.
13. device according to claim 8, which is characterized in that the website information weight statistical module, comprising:
Website information the first merger submodule, for believing the network address recorded under each search term in search history behavior record Breath, according to the theme merger of content corresponding to website information；
Website information weight merger submodule, for counting the weight of the website information under each search term after merger.
14. device according to claim 13, which is characterized in that the website information number of clicks statistical module, comprising:
Website information the second merger submodule, for the website information in the browsing behavior historical record to each user, according to The theme merger of content corresponding to website information；
Number of clicks merger submodule, for counting the number of clicks of the website information under each user identifier after merger.
CN201610327707.2A 2016-05-17 2016-05-17 A kind of information recommendation method and device Expired - Fee Related CN105956161B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201610327707.2A CN105956161B (en) 2016-05-17 2016-05-17 A kind of information recommendation method and device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610327707.2A CN105956161B (en) 2016-05-17 2016-05-17 A kind of information recommendation method and device

Publications (2)

Publication Number Publication Date
CN105956161A CN105956161A (en) 2016-09-21
CN105956161B true CN105956161B (en) 2019-07-02

Family

ID=56911932

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610327707.2A Expired - Fee Related CN105956161B (en) 2016-05-17 2016-05-17 A kind of information recommendation method and device

Country Status (1)

Country Link
CN (1) CN105956161B (en)

Families Citing this family (17)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106709756A (en) * 2016-12-08 2017-05-24 北京五八信息技术有限公司 User demand information acquisition method and apparatus thereof
JP6865045B2 (en) * 2017-01-24 2021-04-28 ヤフー株式会社 Search device, search method and search program
CN107330023B (en) * 2017-06-21 2021-02-12 北京百度网讯科技有限公司 Text content recommendation method and device based on attention points
CN107609902A (en) * 2017-08-28 2018-01-19 口碑(上海)信息技术有限公司 The methods of exhibiting and device of a kind of targeted ads
EP3674908A4 (en) 2017-08-31 2020-07-01 Shenzhen Heytap Technology Corp., Ltd. Information recommendation method and related device
CN107861927A (en) * 2017-09-21 2018-03-30 广州视源电子科技股份有限公司 Document annotation method and device, readable storage medium and computer equipment
CN108153857A (en) * 2017-12-22 2018-06-12 北京奇虎科技有限公司 A kind of method and system for being used to be associated network access data processing
CN108109056A (en) * 2018-01-10 2018-06-01 北京思特奇信息技术股份有限公司 A kind of recommendation method and system of commodity
CN110674386B (en) * 2018-06-14 2022-11-01 北京百度网讯科技有限公司 Resource recommendation method, device and storage medium
CN108876540A (en) * 2018-06-20 2018-11-23 中译语通科技（青岛）有限公司 A kind of closed loop Tourism Marketing system based on big data and across language technology
CN109214418A (en) * 2018-07-25 2019-01-15 百度在线网络技术（北京）有限公司 The method for digging and device, computer equipment and readable medium that user is intended to
CN109272360B (en) * 2018-09-28 2021-09-10 有米科技股份有限公司 Intelligent advertisement recommendation method, system and device
CN109658206B (en) * 2019-01-14 2022-07-26 京东方科技集团股份有限公司 Information recommendation method and device
CN109788344A (en) * 2019-01-30 2019-05-21 四川省有线广播电视网络股份有限公司 Intelligent sound pop-up additional information launches design method
CN111582901B (en) * 2019-02-19 2024-09-20 北京沃东天骏信息技术有限公司 Quantification method, device and system for showing link effect and storage medium
US11720921B2 (en) 2020-08-13 2023-08-08 Kochava Inc. Visual indication presentation and interaction processing systems and methods
CN114693414B (en) * 2022-06-02 2022-11-04 深圳星坊科技有限公司 Store item recommendation method and device, computer equipment and storage medium

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101493832A (en) * 2009-03-06 2009-07-29 辽宁般若网络科技有限公司 Website content combine recommendation system and method
CN102411596A (en) * 2010-09-21 2012-04-11 阿里巴巴集团控股有限公司 Information recommendation method and system
CN102799662A (en) * 2012-07-10 2012-11-28 北京奇虎科技有限公司 Method, device and system for recommending website
CN103118111A (en) * 2013-01-31 2013-05-22 北京百分点信息科技有限公司 Information push method based on data from a plurality of data interaction centers
CN105095279A (en) * 2014-05-13 2015-11-25 深圳市腾讯计算机系统有限公司 File recommendation method and apparatus

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101493832A (en) * 2009-03-06 2009-07-29 辽宁般若网络科技有限公司 Website content combine recommendation system and method
CN102411596A (en) * 2010-09-21 2012-04-11 阿里巴巴集团控股有限公司 Information recommendation method and system
CN102799662A (en) * 2012-07-10 2012-11-28 北京奇虎科技有限公司 Method, device and system for recommending website
CN103118111A (en) * 2013-01-31 2013-05-22 北京百分点信息科技有限公司 Information push method based on data from a plurality of data interaction centers
CN105095279A (en) * 2014-05-13 2015-11-25 深圳市腾讯计算机系统有限公司 File recommendation method and apparatus

Also Published As

Publication number Publication date
CN105956161A (en) 2016-09-21

Similar Documents

Publication Publication Date Title
CN105956161B (en) A kind of information recommendation method and device
CN104217031B (en) A kind of method and apparatus that user's classification is carried out according to server search daily record data
US8250066B2 (en) Search results ranking method and system
US20110238723A1 (en) Systems and methods for web decoding
CN104217030B (en) A kind of method and apparatus that user's classification is carried out according to server search daily record data
US20120143883A1 (en) Ranking product information
CN104715069B (en) It is a kind of to search for the treating method and apparatus for recommending word
CN102831218B (en) Method and device for determining data in thermodynamic chart
WO2013192101A1 (en) Ranking search results based on click through rates
CN105512180B (en) A kind of search recommended method and device
CN107832468A (en) Demand recognition methods and device
WO2008141295A1 (en) Keyword generation system and method for online activity
CN109388548A (en) Method and apparatus for generating information
CN103970796A (en) Inquiry preference ordering method and device
US9330071B1 (en) Tag merging
CN110175895A (en) A kind of item recommendation method and device
JP6437537B2 (en) Embeddable media content search widget
CN104965863B (en) A kind of clustering objects method and apparatus
CN108197244A (en) It is a kind of to search for the method for pushing and device for recommending word
CN106656741A (en) Information push method and system
US9619705B1 (en) Object identification in visual media
US10643178B1 (en) Asynchronous real-time procurement system
CN104281629B (en) The method, apparatus and client device of picture are extracted from webpage
CN111444447A (en) Content recommendation page display method and device
CN109241455A (en) A kind of methods of exhibiting and device of recommended

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
TR01 Transfer of patent right

Effective date of registration: 20220727

Address after: Room 801, 8th floor, No. 104, floors 1-19, building 2, yard 6, Jiuxianqiao Road, Chaoyang District, Beijing 100015

Patentee after: BEIJING QIHOO TECHNOLOGY Co.,Ltd.

Address before: 100088 room 112, block D, 28 new street, new street, Xicheng District, Beijing (Desheng Park)

Patentee before: BEIJING QIHOO TECHNOLOGY Co.,Ltd.

Patentee before: Qizhi software (Beijing) Co.,Ltd.

TR01 Transfer of patent right
CF01 Termination of patent right due to non-payment of annual fee

Granted publication date: 20190702

CF01 Termination of patent right due to non-payment of annual fee