CN102402766B - A kind of user interest modeling method based on web page browsing - Google Patents
A kind of user interest modeling method based on web page browsing Download PDFInfo
- Publication number
- CN102402766B CN102402766B CN201110447908.3A CN201110447908A CN102402766B CN 102402766 B CN102402766 B CN 102402766B CN 201110447908 A CN201110447908 A CN 201110447908A CN 102402766 B CN102402766 B CN 102402766B
- Authority
- CN
- China
- Prior art keywords
- interest
- user
- point
- represent
- commodity
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Information Transfer Between Computers (AREA)
Abstract
The invention discloses a kind of user interest modeling method based on web page browsing, mainly include the following steps that：Using the afterbody catalogue of goods catalogue level as user point of interest；Set up the user interest model for including three dimensions；Wherein, space interest dimension P is used for the tendentiousness for describing user interest；Time interest dimension T is used for the Long-term Interest and short-term interest for describing user；Interest transfer dimension C is used to describe correlation of the user between different points of interest.
Description
Technical field
The present invention relates to field of computer technology, more particularly to the user interest modeling method based on web page browsing.
Background technology
In recent years, with the progress of computer and network technology, ecommerce is developed rapidly.User passes through network
All kinds of commodity can be purchased.In order to help client to find suitable commodity as early as possible, and also to product promotion is done, electronics business
Be engaged in website all can an integrated commending system automatically for client generate commercial product recommending.In commercial product recommending system, in order to be able to accurate
Ground understands the interest characteristics of user more targetedly to do commercial product recommending, it is necessary to the interest founding mathematical models of user.
At present, most commending system is all laid particular emphasis on is simple business according to the purchasing history of user with history is browsed
Product are recommended, and do not set up its interest model for unique user.Its shortcoming is apparent, on the one hand, if user has bought
Certain commodity, this in a short time user the demand of such commodity just have dropped, if now still recommending such commodity, reach
Less than the purpose of product promotion；On the other hand, according only to browse commercial product recommending that history does then only user is confined to it is a few
Among individual commodity, it is impossible to deeply excavate other commodity that user not yet browses.Done in the case where lacking perfect user interest model
Commercial product recommending, it is not fully up to expectations in effect.
The content of the invention
In view of this, a kind of mathematical modeling that can reflect user interest in all its bearings is highly beneficial.
In order to solve the above problems, the invention provides a kind of user interest modeling method based on web page browsing, its skill
Art scheme comprises the following steps：
1. according to the commodity classification bibliographic structure of e-commerce website, it regard the afterbody catalogue of TOC level as user
Point of interest, point of interest includes all commodity pages under the catalogue；
2. the interest model of user includes three dimensions：Space interest dimension P, time interest dimension T, interest transfer dimension
C, by user in the record that browses of website, calculates above three dimension；
3. space interest dimension P is used for the tendentiousness for describing user interest, it is expressed as：
P={ (W0, IP0), (W1, IP1) ..., (Wn, IPn)}
Wherein, WiRepresent user's Access Interest point i total degree, IPiUser is described to each commodity in point of interest i
Interest vector, its expression formula is：
IPi=[w0, w1..., wm]
Wherein, wjRepresent interest-degree of the user to commodity j in point of interest；
4. time interest dimension T is used for the Long-term Interest and short-term interest for describing user, it is expressed as：
T={ (LT0, ST0), (LT1, ST1) ..., (LTn, STn)}
Wherein, LTiRepresent total line duration that user is accessed point of interest i, STiRepresent the nearest Access Interest point of user
Density；
5. interest transfer dimension C is used to describe correlation of the user between different points of interest, its expression formula is：
Ix, IyRepresent two different point of interest x and y, P (Ix, Iy) session of user is represented while comprising point of interest x and y
Probability, P (Ix) represent the probability that the session of user includes point of interest x, P (Iy) to represent the session of user general comprising point of interest y
Rate, C (Ix, Iy) point of interest x and y correlation is reflected, value is more than 1 and represents positive correlation, and value is less than 1 and represents negatively correlated, and value is equal to
1 represents uncorrelated.
Wherein, n+1 represents the total number of point of interest, and m+1 represents the total number of the commodity in each point of interest, and i is integer,
And 0≤i≤n, j are integer, and 0≤j≤m.
The present invention can also pass through following methods enhancing modeling effect：
Interest-degree w of the user to commodity in point of interestjComputational methods it is as follows：
Wherein, CNT (j) represents access times of the user to commodity j in point of interest, and PU represents the affiliated point of interest institute of the commodity
The total number of users possessed, IPUjRepresent the total number of users that the commodity are possessed.
The calculation formula for total line duration that user is accessed point of interest i is：
Wherein, a represents the number of times of user's access, tkRepresent kth time and access the spent time.
User is to the calculation formula of point of interest i nearest Access Interest density：
Wherein, θ represents a fixed time interval, tkθRepresent since the current time in k θ time range forward,
User's Access Interest point i total time, b is a parameter, the time span scope for setting the calculating of Access Interest density.
The above-mentioned user interest modeling method based on web page browsing opens one kind and obtained by analyzing user's internet behavior
The approach of user interest profile.It from 3 not ipsilateral describe user interest profile：Time, space, interest transfer.Compare
It is presently recommended that being based only on the simple model that user browses record and purchaser record employed in system, the present invention can be more
The actual interest orientation being close to the users.
Brief description of the drawings
Fig. 1 shows user interest modeling method flow；
Embodiment
The present invention is described in further detail with embodiment below in conjunction with the accompanying drawings：
Fig. 1 shows user interest modeling method flow, mainly includes：
Step S100, interest point extraction.Using the afterbody catalogue of goods catalogue level as user point of interest, one
Point of interest includes all commodity pages under the catalogue；
Step S101, user's browse state record.Including the point of interest residing for user's browsing pages, the trade name of the page
Claim, user click frequency, user's residence time, the number of users of current commodity updates；
Step S102, space interest dimension is calculated.Space interest dimension P expression formula is：
P={ (W0, IP0), (W1, IP1) ..., (Wn, IPn)}
Wherein, WiRepresent user's Access Interest point i total degree, IPiUser is described to each commodity in point of interest i
Interest vector, its expression formula is：
IPi=[w0, w1..., wm]
Wherein, wjInterest-degree of the user to commodity j in point of interest is represented, its computational methods is：
Wherein, CNT (j) represents access times of the user to commodity j in point of interest, and PU represents the affiliated point of interest institute of the commodity
The total number of users possessed, IPUjRepresent the total number of users that the commodity are possessed.
Step S103, time interest dimension is calculated.Time, interest dimension T was expressed as：
T={ (LT0, ST0), (LT1, ST1) ..., (LTn, STn)}
Wherein, LTiRepresent total line duration that user is accessed point of interest i, STiRepresent the nearest Access Interest point of user
Density, the calculation formula for total line duration that user is accessed point of interest i is：
Wherein, a represents the number of times of user's access, tkRepresent kth time and access the spent time.
User is to the calculation formula of point of interest i nearest Access Interest density：
Wherein, θ represents a fixed time interval, tkθRepresent since the current time in k θ time range forward,
User's Access Interest point i total time, b is a parameter, the time span scope for setting the calculating of Access Interest density.
Step S104, interest transfer dimension is calculated.Interest transfer dimension C expression formula is：
Ix, IyRepresent two different point of interest x and y, P (Ix, Iy) session of user is represented while comprising point of interest x and y
Probability, P (Ix) represent the probability that the session of user includes point of interest x, P (Iy) to represent the session of user general comprising point of interest y
Rate, C (Ix, Iy) point of interest x and y correlation is reflected, value is more than 1 and represents positive correlation, and value is less than 1 and represents negatively correlated, and value is equal to
1 represents uncorrelated.
Wherein, n+1 represents the total number of point of interest, and m+1 represents the total number of the commodity in each point of interest, and i is integer,
And 0≤i≤n, j are integer, and 0≤j≤m.
Claims (4)
1. a kind of user interest modeling method based on web page browsing, it is characterised in that comprise the following steps：
1) according to the commodity classification bibliographic structure of e-commerce website, it regard the afterbody catalogue of TOC level as the emerging of user
Interesting, a point of interest includes all commodity pages under the catalogue；
2) interest model of user includes three dimensions：Space interest dimension P, time interest dimension T, interest transfer dimension C, lead to
Browse record of the user in website is crossed, above three dimension is calculated；
3) space interest dimension P is used for the tendentiousness for describing user interest, and it is expressed as：
P={ (W0, IP0), (W1, IP1) ..., (Wn, IPn)}
Wherein, WiRepresent user's Access Interest point i total degree, IPiDescribe user in point of interest i each commodity it is emerging
Interesting vector, its expression formula is：
IPi=[w0, w1..., wm]
Wherein, wjRepresent interest-degree of the user to commodity j in point of interest；
4) time interest dimension T is used for the Long-term Interest and short-term interest for describing user, and it is expressed as：
T={ (LT0, ST0), (LT1, ST1) ..., (LTn, STn)}
Wherein, LTiRepresent total line duration that user is accessed point of interest i, STiRepresent the density of the nearest Access Interest point of user；
5) interest transfer dimension C is used to describe correlation of the user between different points of interest, and its expression formula is：
Ix, IyRepresent two different point of interest x and y, P (Ix, Iy) session of user is represented while general comprising point of interest x and y
Rate, P (Ix) represent the probability that the session of user includes point of interest x, P (Iy) represent the probability that the session of user includes point of interest y, C
(Ix, Iy) point of interest x and y correlation is reflected, value is more than 1 and represents positive correlation, and value is less than 1 and represents negatively correlated, and value is equal to 1 table
Show uncorrelated；
Wherein, n+1 represents the total number of point of interest, and m+1 represents the total number of the commodity in each point of interest, and i is integer, and 0
≤ i≤n, j are integer, and 0≤j≤m.
2. according to the method described in claim 1, it is characterised in that interest-degree w of the user to commodity in point of interestjCalculating side
Method is as follows：
Wherein, CNT (j) represents access times of the user to commodity j in point of interest, and PU represents the affiliated point of interest of the commodity and possessed
Total number of users, IPUjRepresent the total number of users that the commodity are possessed.
3. according to the method described in claim 1, it is characterised in that the calculating for total line duration that user is accessed point of interest i
Formula is：
Wherein, a represents the number of times of user's access, tkRepresent kth time and access the spent time.
4. the method according to claim 1 or 3, it is characterised in that user is to point of interest i nearest Access Interest density
Calculation formula is：
Wherein, θ represents a fixed time interval, tkθRepresent since the current time in k θ time range forward, user
Access Interest point i total time, b is a parameter, the time span scope for setting the calculating of Access Interest density.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110447908.3A CN102402766B (en) | 2011-12-27 | 2011-12-27 | A kind of user interest modeling method based on web page browsing |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110447908.3A CN102402766B (en) | 2011-12-27 | 2011-12-27 | A kind of user interest modeling method based on web page browsing |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102402766A CN102402766A (en) | 2012-04-04 |
| CN102402766B true CN102402766B (en) | 2017-07-28 |
Family
ID=45884946
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201110447908.3A Active CN102402766B (en) | 2011-12-27 | 2011-12-27 | A kind of user interest modeling method based on web page browsing |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102402766B (en) |
Families Citing this family (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102760163B (en) * | 2012-06-12 | 2015-04-29 | 北京奇虎科技有限公司 | Personalized recommendation method and device of characteristic information |
| CN103235826B (en) * | 2013-05-06 | 2016-06-15 | 上海河广信息科技有限公司 | A kind of control method of time window |
| CN104636406B (en) * | 2013-11-15 | 2018-05-18 | 华为技术有限公司 | A kind of method and apparatus according to user behavior pushed information |
| CN103780625B (en) * | 2014-01-26 | 2017-07-04 | 北京搜狗科技发展有限公司 | User interest finds method and apparatus |
| CN104462560B (en) * | 2014-12-25 | 2018-01-05 | 广东电子工业研究院有限公司 | A kind of recommendation method of personalized recommendation system |
| CN106339502A (en) * | 2016-09-18 | 2017-01-18 | 电子科技大学 | Modeling recommendation method based on user behavior data fragmentation cluster |
| CN109727047A (en) * | 2017-10-30 | 2019-05-07 | 北京京东尚科信息技术有限公司 | A kind of method and apparatus, data recommendation method and the device of determining data correlation degree |
| CN107844598A (en) * | 2017-11-22 | 2018-03-27 | 广州优视网络科技有限公司 | Content recommendation method, device and computer equipment |
| CN110245299B (en) * | 2019-06-19 | 2022-02-08 | 中国人民解放军国防科技大学 | Sequence recommendation method and system based on dynamic interaction attention mechanism |
| CN111125537B (en) * | 2019-12-31 | 2020-12-22 | 中国计量大学 | Session recommendation method based on graph representation |
| CN112818224B (en) * | 2021-01-26 | 2024-02-20 | 北京百度网讯科技有限公司 | Information recommendation method and device, electronic equipment and readable storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101520878A (en) * | 2009-04-03 | 2009-09-02 | 华为技术有限公司 | Method, device and system for pushing advertisements to users |
| CN101826114A (en) * | 2010-05-26 | 2010-09-08 | 南京大学 | Multi Markov chain-based content recommendation method |
| CN101923545A (en) * | 2009-06-15 | 2010-12-22 | 北京百分通联传媒技术有限公司 | Method for recommending personalized information |
| CN102215300A (en) * | 2011-05-24 | 2011-10-12 | 中国联合网络通信集团有限公司 | Telecommunication service recommendation method and system |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6981040B1 (en) * | 1999-12-28 | 2005-12-27 | Utopy, Inc. | Automatic, personalized online information and product services |
| US8432919B2 (en) * | 2009-02-25 | 2013-04-30 | Cisco Technology, Inc. | Data stream classification |
-
2011
- 2011-12-27 CN CN201110447908.3A patent/CN102402766B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101520878A (en) * | 2009-04-03 | 2009-09-02 | 华为技术有限公司 | Method, device and system for pushing advertisements to users |
| CN101923545A (en) * | 2009-06-15 | 2010-12-22 | 北京百分通联传媒技术有限公司 | Method for recommending personalized information |
| CN101826114A (en) * | 2010-05-26 | 2010-09-08 | 南京大学 | Multi Markov chain-based content recommendation method |
| CN102215300A (en) * | 2011-05-24 | 2011-10-12 | 中国联合网络通信集团有限公司 | Telecommunication service recommendation method and system |
Non-Patent Citations (4)
| Title |
|---|
| 基于浏览行为量化计算的用户兴趣网页分析;周智;《中国优秀硕士学位论文全文数据库科技信息辑》;20091215(第12期);第I139-133页 * |
| 基于用户浏览行为分析的用户兴趣获取;尹春晖等;《计算机技术与发展》;20080531;第18卷(第5期);第37-39页 * |
| 基于网页浏览的兴趣行为研究;米雪等;《电脑知识与技术》;20101231;第6卷(第34期);第9834-9836页 * |
| 基于隐马尔科夫模型的浏览兴趣预测;孙秀娟等;《科技导报》;20090930;第27卷(第18期);第75-77页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102402766A (en) | 2012-04-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102402766B (en) | A kind of user interest modeling method based on web page browsing | |
| CN103412948B (en) | The Method of Commodity Recommendation and system of collaborative filtering based on cluster | |
| CN102542489B (en) | One kind is based on the associated recommendation method of user interest | |
| Ma et al. | Recommender systems with social regularization | |
| CN106372249B (en) | A kind of clicking rate predictor method, device and electronic equipment | |
| CN107016030B (en) | Keyword estimation value feedback method and system | |
| US20170206361A1 (en) | Application recommendation method and application recommendation apparatus | |
| US8666834B2 (en) | Item recommendation system, item recommendation method and program | |
| US20140108200A1 (en) | Method and system for recommending search phrases | |
| CN107464132B (en) | Similar user mining method and device and electronic equipment | |
| CN102542046A (en) | Book recommendation method based on book contents | |
| WO2018121700A1 (en) | Method and device for recommending application information based on installed application, terminal device, and storage medium | |
| CN102073717A (en) | Home page recommending method for orienting vertical e-commerce website | |
| CN103399858A (en) | Socialization collaborative filtering recommendation method based on trust | |
| CN105005917A (en) | Universal method for correlating single items of different e-commerce websites | |
| CN103106600A (en) | Commodity information push system and commodity information push method | |
| CN103942193B (en) | A kind of method and device of information push | |
| EP2895969A2 (en) | Dynamic data acquisition method and system | |
| CN105023178B (en) | A kind of electronic commerce recommending method based on ontology | |
| WO2016107455A1 (en) | Information matching processing method and apparatus | |
| CN103870575A (en) | Method and device for extracting domain keywords | |
| US8880518B2 (en) | Ranking products using purchase day based time windows | |
| CN103150667A (en) | Individualization recommending method based on mechanical structure | |
| CN105894310A (en) | Personalized recommendation method | |
| Xiao et al. | An item based collaborative filtering system combined with genetic algorithms using rating behavior |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C41 | Transfer of patent application or patent right or utility model | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20160921 Address after: East Building 11, 100195 Beijing city Haidian District xingshikou Road No. 65 west Shan creative garden district 1-4 four layer of 1-4 layer Applicant after: Beijing Jingdong Shangke Information Technology Co., Ltd. Address before: 201203 Shanghai city Pudong New Area Zu Road No. 295 Room 102 Applicant before: Niuhai Information Technology (Shanghai) Co., Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20201106 Address after: No.8-6, Putou South Road, Haicang District, Xiamen City, Fujian Province Patentee after: Xiamen xinjianfu e-commerce Co., Ltd Address before: East Building 11, 100195 Beijing city Haidian District xingshikou Road No. 65 west Shan creative garden district 1-4 four layer of 1-4 layer Patentee before: BEIJING JINGDONG SHANGKE INFORMATION TECHNOLOGY Co.,Ltd. |
|
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20210512 Address after: 361000 No.8, Putou South Road, Haicang District, Xiamen City, Fujian Province Patentee after: Xiamen Jianfu Chain Management Co.,Ltd. Address before: No.8-6, Putou South Road, Haicang District, Xiamen City, Fujian Province 361022 Patentee before: Xiamen xinjianfu e-commerce Co., Ltd |