CN101853463A - Collaborative filtering recommending method and system based on client characteristics - Google Patents
Collaborative filtering recommending method and system based on client characteristics Download PDFInfo
- Publication number
- CN101853463A CN101853463A CN200910080946A CN200910080946A CN101853463A CN 101853463 A CN101853463 A CN 101853463A CN 200910080946 A CN200910080946 A CN 200910080946A CN 200910080946 A CN200910080946 A CN 200910080946A CN 101853463 A CN101853463 A CN 101853463A
- Authority
- CN
- China
- Prior art keywords
- client
- user
- recommendation
- collaborative filtering
- algorithm
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses collaborative filtering recommending method and system based on client characteristics. The method comprises the following steps of: obtaining user information and carrying out information expression, neighbor formation and recommendation generation. Based on a most adjacent collaborative filtering algorithm, the method comprises algorithm name, algorithm input and algorithm output, and also comprises algorithm analysis. The invention also discloses a collaborative filtering recommending system based on client characteristic forecasting, which comprises data loading, recommendation engine as well as client mail notification and recommendation daemon process.
Description
Technical field
The present invention relates to technical field of the computer network, relate in particular to a kind of collaborative filtering recommending method and system based on client characteristics.
Background technology
Along with popularizing and Development of E-business of internet, an important research content of commending system especially personalized recommendation becoming gradually e-commerce technology obtains researcher's concern more and more.At present, nearly all large-scale e-commerce system, as Amazon, CDNOW, eBay, bookstore etc. has all used various forms of commending systems to some extent in the Dangdang.com.
Commending system is exactly to utilize the system that statistics and knowledge discovering technologies solve with the target customer provides the commercial product recommending problem when mutual.It provides merchandise news and suggestion to the client in e-commerce system, help the client to determine to buy which kind of commodity, and the pseudo sale personnel finish the process of purchase to the lead referral commodity.Which kind of commodity commending system recommends is to analyze generation on the demographics of purchase situation, client at the whole commodity of e-commerce website or the historical record that the client is bought.
The recommendation information that early stage commending system provides is to provide identical information to all users, as the ranking list of merchandise sales, promotion mix etc., these recommendation informations are not to recommend at the individual consumer, and this class commending system exists not enough aspect personalized recommendation.With work marketing ideas and Development of E-business, for adapting to the needs that customer requirement in the actual business environment obtains man-to-man service, the commending system that is applied to e-commerce website is gradually to developing for the client provides this direction of recommendation service that meets its individual demand in real time.
Under this background, the further raising information service quality that appears as of collaborative filtering provides a new thinking.Collaborative filtering, its basic thought is very directly perceived, and in daily life, people tend to make some selections (shopping, reading, music or the like) according to relatives and friends' recommendation.The collaborative filtering system applies to this thought in the network information service (information recommendation) exactly, based on other users the evaluation of a certain information is recommended to a certain user.
Further expansion along with the e-commerce system scale, number of users and project data sharply increase, the extremely sparse property that causes user's score data, under the extremely sparse situation of user's score data, all there is drawback separately in the whole bag of tricks of tradition similarity measurement, the feasible targeted customer's who calculates nearest-neighbors is inaccurate, and the recommendation quality of commending system sharply descends.The more important thing is that these algorithms only are suitable under the prerequisite of record there being the user to mark, at the e-commerce system that does not have the user to mark, can't be in Technologies of Recommendation System in E-Commerce widespread use collaborative filtering technology.
Summary of the invention
The present invention has proposed a kind of collaborative filtering recommending method based on the client characteristics prediction for solving the problems of the technologies described above, and at first by the consumer behavior eigenvalue matrix of computing client, goes out client's nearest-neighbors again in conjunction with client's purchaser record COMPREHENSIVE CALCULATING.It is extremely sparse and do not have the deficiency that traditional method for measuring similarity exists under the marking situation of client to commodity that this method can solve user's score data effectively, improves the recommendation quality of commending system.
The invention provides a kind of collaborative filtering recommending method, mainly may further comprise the steps based on client characteristics:
1, obtain user profile and carry out information representation, user profile obtain the evaluation of main acquisition user to given information (finger of broad sense, narrow sense can refer to commodity that businessman provides etc.).Wherein: evaluation is divided into explicit evaluation and implicit expression is estimated two kinds.
Explicit evaluation needs the user to express own approval degree to a certain information consciously, generally represents the degree liked with round values, as GroupLens, Ringo etc.The collaborative filtering system provides an information list to new user, require the user that all or part of information is wherein estimated, after system obtains these initial informations of user, just the user can be joined in the user library, along with the user constantly uses the collaborative filtering recommending system, user profile constantly accumulates, thereby provides the recommendation that more and more meets its individual demand for the user.
The implicit expression evaluation is wished to obtain user profile from user's behavior, and the research of having done at present has by analysis user shopping online record, the length of reading the article time, the data recording such as connection number of times of URL obtains.Just present, implicit expression is estimated bigger to the difficulty of data analysis, accuracy and validity remain further to be improved, so the research in collaborative filtering field mainly is evaluated as the master with explicit, on development potentiality, along with improving constantly of the analysis of implicit expression evaluating data, processing power, the commending system of utilization implicit expression evaluation will account for increasing proportion.
The description of user profile is mainly finished in information representation, generally uses matrix representation of user, and item is represented commodity (or given information, as: film, books etc.).
2, neighbours form, and neighbours form the identification of mainly finishing targeted customer's nearest-neighbors (or similar users).
The core that collaborative filtering is realized in commending system is exactly to be one to need the current goal user of recommendation service to seek its most similar " nearest-neighbors " collection, promptly, produce " neighbours " set N={N1, N2 according to the big minispread of similarity to a user u, N,, u does not belong to N, from N1, to Nt, sim (u, N) (value of expression user u and neighbours user's similarity size) arrangement from big to small.
3, produce recommendation, generation is recommended mainly to solve from nearest-neighbors information and is obtained the prediction of targeted customer's interest-degree or the recommendation of information.
The prediction of user interest degree can calculate by formula, provides optimal N item (Top-N) to recommend and can realize by two kinds of different technology to the user: recommend and based on the recommendation of correlation rule for the most frequent.
Recommend for the most frequent: each user i that statistics nearest-neighbors user concentrates is to the interest-degree of different item, can weigh with the size of access frequency or interest-degree, get wherein N and come the foremost, and the item that the active user also not have to buy or estimate is recommended to collect as Top-N.
Recommendation based on correlation rule: the top-N that this recommendation is based on the correlation rule technology recommends.Different is, what use when producing correlation rule is not whole data set but nearest-neighbors data set, and this has reduced calculated amount to a certain extent.In practice, only considering that neighbours collect the formation of the correlation rule that data produce may be abundant inadequately, thereby influences the accuracy of last recommendation results, therefore carry out Top-N recommend to select for use often before a kind of method.
Described method comprises collaborative filtering based on arest neighbors as the basis, and comprising: the algorithm title, algorithm input and algorithm are exported.Wherein:
Algorithm title: based on the collaborative filtering of client characteristics
Algorithm input: client's purchase inventory records
Algorithm output: the commodity collection is bought in each client's recommendation, according to priority ordered
Described method also comprises Algorithm Analysis, collaborative filtering computation schema based on client characteristics remains based on correlativity, mainly calculates in user's space, and complexity has reached O (n2), wherein n is the number of users of E-business applications, and calculated amount is relatively large.But consider generally relatively stable within 1-2 month under the typical electronic business environment to the similarity between the user, this stability makes our similarity between the computing client in advance, for example wherein a kind of possible method is to calculate some neighbours (similar client) of all clients, utilize the table storage then, so that with timely, fast query then.The L of searched targets user Ui similar redeeming oneself through calculating good similar users in advance at first when algorithm is realized checked the commodity collection at this L similar client's purchaser record then, produces last recommendation according to the multiplicity of commodity.
The present invention also provides a kind of collaborative filtering recommending system based on the client characteristics prediction, comprising: Data Loading, recommended engine, client's mail notification and recommendation finger daemon.Wherein:
Data Loading is finished the conversion from the operation system database of website to the commending system desired data, main function: the extraction of data, the loading of the cleaning of data and conversion and data
Recommended engine is the core of system, finishes the calculating to user's Recommendations, has realized two kinds of algorithms at present: Apriori algorithm and the collaborative filtering based on client characteristics disclosed by the invention.
Client's mail notification, the recommendation of being responsible for recommended engine is calculated sends to client's mailbox: install the letter body and send to client's mail by smtp server by configurable mail template group.
Recommending finger daemon is the finger daemon of total system, and the startup of finishing recommended engine and outgoing mail server such as stops at scheduling and mistake is restarted.
The systematic parameter of total system, recommendation calculating parameter all are placed in the xml file, can be configured easily.
Description of drawings
Fig. 1 is the realization synoptic diagram of collaborative filtering in the inventive method;
Fig. 2 is the collaborative filtering recommending method embodiment schematic flow sheet that the present invention is based on client characteristics;
Fig. 3 is that user profile is represented synoptic diagram in the inventive method;
Fig. 4 is that neighbours form synoptic diagram in the inventive method;
Fig. 5 is the collaborative filtering synoptic diagram that the present invention is based on arest neighbors;
Fig. 6 is a collaborative filtering synoptic diagram of the present invention;
Fig. 7 is the collaborative filtering recommending system architecture synoptic diagram that the present invention is based on the client characteristics prediction.
Embodiment
In order to make those skilled in the art person understand the scheme of the embodiment of the invention better, the embodiment of the invention is described in further detail below in conjunction with drawings and embodiments.
With reference to Fig. 1 is the realization synoptic diagram of collaborative filtering in the inventive method, and describing intuitively for one of collaborative filtering is that user and data entries are constituted a matrix, i.e. the interest matrix of user-data entries.Existing value is the evaluation of user to the corresponding information clauses and subclauses in the matrix, the prediction that unknown-value needs system to provide just.The process of collaborative filtering is predicted unknown-value (process of filling a vacancy) according to given value exactly.The applied algorithm of collaborative filtering system is exactly the rule that this process of filling a vacancy is followed, and regular and actual rule meets more, and the unknown-value of prediction is just accurate more, and the effect of information filtering will be good more.It should be noted that in the reality that such matrix is a very sparse matrix.Because each user general only can in all data entries seldom a part evaluation is arranged.
With reference to Fig. 2 is the collaborative filtering recommending method schematic flow sheet that the present invention is based on client characteristics, a collaborative filtering system, its target is to allow the user cooperate together, find their interested content from a large amount of client's consumption features, obtain user profile, similarity between the analysis user forms neighbours' collection of current goal user, thereby recommends according to the information of " neighbours ".
With reference to Fig. 3 is that user profile is represented synoptic diagram in the inventive method, and one of the user of a m * n estimates matrix R, and m shows number of users, and n has shown item number, r
I, jBe the evaluation of estimate (or assessed value) of i user to the j item, evaluation of estimate is relevant with the content of item, if item is the product in the ecommerce, whether then represent user's ordering products, for example 1 expression is ordered, 0 expression not have to order: or how high the expression user have to its interest-degree, can represent different grades by different values, as 1-7 etc.If be the Web document, then expression whether browse or the user to its interest-degree.
With reference to Fig. 4 is that neighbours form synoptic diagram in the inventive method, mainly may further comprise the steps:
Step 401 is calculated similarity between current usefulness 0 and other user according to certain or certain the several correlation properties between the user.
Step 402 according to the range size of neighbours user's collection of setting or the standard value of correlation properties, is carried out neighbours user's selection.For example establish neighbours user's number k=5, be selected as neighbours for the individual user recently of the k=5 at center with point 0.
With reference to Fig. 5 is the collaborative filtering synoptic diagram that the present invention is based on arest neighbors, based on collaborative filtering based on arest neighbors, the first step at original algorithm is represented in the data, have only client-matrix of item to be used for searching the neighbours that calculate each client, will add client-client characteristics value matrix in improved algorithm calculates, by calculating the similarity of client and client in these two matrixes respectively, these two similaritys of weighting then, obtain two clients' final similarity, according to the subsequent step of collaborative filtering, finally generate recommendation results again.
With reference to Fig. 6 is collaborative filtering synoptic diagram of the present invention, and algorithm flow comprises step:
Step 601, some consumer behavior eigenwerts of each client are added up and calculated to inquiry client's historical purchaser record, and the result is deposited in client-eigenvalue matrix C, and for the element value Ci in the matrix, j=c represents that j the eigenwert of user i is c
Step 602, inquiry client's historical purchaser record gathers to become and buys merchandise classification quantity generation client-purchase categorical measure matrix P, and for the element value Pi in the matrix, j=p represents that user i has bought p the commodity that classification is j
Step 603 is calculated each client's n the most similar client (neighbours), and the similarity between the client is defined as follows:
Sim(Ui，Uj)＝w1*Sim_c(Ci，Cj)+w2*Sim_p(Pi，Pj)
Sim_c (Ci, Cj) similarity of client i and client j among expression client-eigenvalue matrix C wherein
Sim_p (w2 is the weight of two similaritys for Ci, Cj) the similarity w1 of client i and client j among expression client-purchase classification matrix P, and client's similarity is calculated respectively according to the cosine relevance formula among the general w1+w2=1, two matrixes:
Step 604, to each client Ui, search n neighbours that obtain with his degree of correlation maximum:
Neighbors(Ui)＝Top-N(Sim(Ui，U))
Step 605 for each client Ui, calculates last recommendation by the 4 neighbours clients that obtain:
Recommends(Ui)＝E(Perchase(Neighbors(Ui)))
With reference to Fig. 7 is the collaborative filtering recommending system architecture synoptic diagram that the present invention is based on the client characteristics prediction, comprising: Data Loading, recommended engine, client's mail notification and recommendation finger daemon.
More than the invention process is described in detail, used embodiment herein the present invention set forth, the explanation of above embodiment just is used for help understanding method and system of the present invention; Simultaneously, for one of ordinary skill in the art, according to thought of the present invention, the part that all can change in specific embodiments and applications, in sum, this description should not be construed as limitation of the present invention.
Claims (8)
1. the collaborative filtering recommending method based on client characteristics is characterized in that, comprising: obtain user profile and carry out information representation, neighbours form and produce and recommend.Wherein:
Obtain user profile, mainly the evaluation of given information (finger of broad sense, narrow sense can refer to commodity that businessman provides etc.) is obtained by the user.
Neighbours form, and mainly finish the identification of targeted customer's nearest-neighbors (or similar users).
Generation is recommended, and mainly solves from nearest-neighbors information to obtain the prediction of targeted customer's interest-degree or the recommendation of information.
2. method according to claim 1 is characterized in that, described method comprises collaborative filtering based on arest neighbors as the basis, and comprising: the algorithm title, algorithm input and algorithm are exported.Wherein:
Algorithm title: based on the collaborative filtering of client characteristics
Algorithm input: client's purchase inventory records
Algorithm output: the commodity collection is bought in each client's recommendation, according to priority ordered.
3. method according to claim 1 is characterized in that, described method also comprises Algorithm Analysis.
4. method according to claim 1 is characterized in that, the evaluation of given information is divided into explicit evaluation to described user and implicit expression is estimated two kinds.
5. method according to claim 1 is characterized in that, neighbours form one in the described method needs the current goal user of recommendation service to seek its most similar " nearest-neighbors " collection, that is: to a user u, produce " neighbours " set N={N1, N2 according to the big minispread of similarity, N,, u does not belong to N, from N1, to Nt, sim (u, N) (value of expression user u and neighbours user's similarity size) arrangement from big to small.Recommendation based on correlation rule: the top-N that this recommendation is based on the correlation rule technology recommends.
6. method according to claim 1 is characterized in that, produces in the described method to recommend the top-N that is based on the correlation rule technology to recommend.
7. the collaborative filtering recommending system based on the client characteristics prediction comprises: Data Loading, recommended engine, client's mail notification and recommendation finger daemon.Wherein:
Data Loading is finished the conversion from the operation system database of website to the commending system desired data, main function: the extraction of data, the loading of the cleaning of data and conversion and data.
Recommended engine is the core of system, finishes the calculating to user's Recommendations.
Client's mail notification, the recommendation of being responsible for recommended engine is calculated sends to client's mailbox: install the letter body and send to client's mail by smtp server by configurable mail template group.
Recommending finger daemon is the finger daemon of total system, and the startup of finishing recommended engine and outgoing mail server such as stops at scheduling and mistake is restarted.
The systematic parameter of total system, recommendation calculating parameter all are placed in the xml file, can be configured easily.
8. system according to claim 7 is characterized in that, recommended engine has been realized two kinds of algorithms in the described system: Apriori algorithm and based on the collaborative filtering of client characteristics.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200910080946A CN101853463A (en) | 2009-03-30 | 2009-03-30 | Collaborative filtering recommending method and system based on client characteristics |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200910080946A CN101853463A (en) | 2009-03-30 | 2009-03-30 | Collaborative filtering recommending method and system based on client characteristics |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101853463A true CN101853463A (en) | 2010-10-06 |
Family
ID=42804932
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN200910080946A Pending CN101853463A (en) | 2009-03-30 | 2009-03-30 | Collaborative filtering recommending method and system based on client characteristics |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101853463A (en) |
Cited By (27)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102567900A (en) * | 2011-12-28 | 2012-07-11 | 尚明生 | Method for recommending commodities to customers |
| CN102629360A (en) * | 2012-03-13 | 2012-08-08 | 浙江大学 | Effective dynamic commodity recommendation method and commodity recommendation system |
| CN102750288A (en) * | 2011-04-21 | 2012-10-24 | 中国移动通信集团广东有限公司 | Internet content recommending method and device |
| CN102779131A (en) * | 2011-05-12 | 2012-11-14 | 同济大学 | Collaborative filtering recommending method based on multiple-similarity of users |
| CN102799623A (en) * | 2012-06-21 | 2012-11-28 | 杜小勇 | Information pushing method |
| CN103020855A (en) * | 2012-12-11 | 2013-04-03 | 北京京东世纪贸易有限公司 | Bad commodity distinguishing method and system based on user purchasing behavior |
| CN103218407A (en) * | 2013-03-22 | 2013-07-24 | 南京信通科技有限责任公司 | Recommendation engine based on interest graph |
| CN103345503A (en) * | 2013-07-01 | 2013-10-09 | 杭州万事利丝绸科技有限公司 | Silk product individualized recommendation method based on wavelet network |
| CN103353865A (en) * | 2013-04-25 | 2013-10-16 | 杭州电子科技大学 | Barter electronic trading commodity recommendation method based on position |
| CN103412865A (en) * | 2013-06-07 | 2013-11-27 | 广州唯品会信息科技有限公司 | Method and system for carrying out notification on website projects |
| CN104572880A (en) * | 2014-12-22 | 2015-04-29 | 中国科学院信息工程研究所 | Method and system for realizing concurrent cooperated filtering based on users |
| CN104809626A (en) * | 2015-03-17 | 2015-07-29 | 徐邑江 | Customized commodity recommending method based on user credit assessment |
| CN104899236A (en) * | 2014-11-13 | 2015-09-09 | 深圳市腾讯计算机系统有限公司 | Comment information display method, comment information display device and comment information display system |
| CN105096149A (en) * | 2014-05-09 | 2015-11-25 | 华为软件技术有限公司 | Service content product correlation analysis method and apparatus |
| CN105225135A (en) * | 2015-10-31 | 2016-01-06 | 广州华多网络科技有限公司 | Potentiality customer recognition method and device |
| CN105302880A (en) * | 2015-10-14 | 2016-02-03 | 合一网络技术(北京)有限公司 | Content correlation recommendation method and apparatus |
| CN106547365A (en) * | 2015-09-17 | 2017-03-29 | 阿里巴巴集团控股有限公司 | The method and apparatus of commercial product recommending |
| CN106651546A (en) * | 2017-01-03 | 2017-05-10 | 重庆邮电大学 | Intelligent community oriented electronic commerce information recommendation method |
| CN106777051A (en) * | 2016-12-09 | 2017-05-31 | 重庆邮电大学 | A kind of many feedback collaborative filtering recommending methods based on user's group |
| CN107644036A (en) * | 2016-07-21 | 2018-01-30 | 阿里巴巴集团控股有限公司 | A kind of method, apparatus and system of data object push |
| CN109299349A (en) * | 2018-09-11 | 2019-02-01 | 广州视源电子科技股份有限公司 | Application recommendation method and device, equipment and computer-readable storage medium |
| CN109740914A (en) * | 2018-12-28 | 2019-05-10 | 武汉金融资产交易所有限公司 | A kind of method, storage medium, equipment and system that financial business is assessed, recommended |
| CN109857498A (en) * | 2019-01-09 | 2019-06-07 | 明基智能科技（上海）有限公司 | Intelligent content template recommender system and its method |
| CN109951724A (en) * | 2017-12-20 | 2019-06-28 | 阿里巴巴集团控股有限公司 | Recommended method, main broadcaster's recommended models training method and relevant device is broadcast live |
| CN109995837A (en) * | 2018-01-02 | 2019-07-09 | 中国移动通信有限公司研究院 | A kind of service package recommended method, device and server |
| CN111210253A (en) * | 2019-11-26 | 2020-05-29 | 恒大智慧科技有限公司 | Method, device and storage medium for matching consumer with sales consultant |
| CN111882400A (en) * | 2020-07-31 | 2020-11-03 | 平安国际融资租赁有限公司 | Behavior recognition analysis method and device, computer equipment and readable storage medium |
-
2009
- 2009-03-30 CN CN200910080946A patent/CN101853463A/en active Pending
Cited By (42)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102750288A (en) * | 2011-04-21 | 2012-10-24 | 中国移动通信集团广东有限公司 | Internet content recommending method and device |
| CN102750288B (en) * | 2011-04-21 | 2015-11-11 | 中国移动通信集团广东有限公司 | A kind of internet content recommend method and device |
| CN102779131B (en) * | 2011-05-12 | 2015-02-18 | 同济大学 | Collaborative filtering recommending method based on multiple-similarity of users |
| CN102779131A (en) * | 2011-05-12 | 2012-11-14 | 同济大学 | Collaborative filtering recommending method based on multiple-similarity of users |
| CN102567900A (en) * | 2011-12-28 | 2012-07-11 | 尚明生 | Method for recommending commodities to customers |
| CN102629360A (en) * | 2012-03-13 | 2012-08-08 | 浙江大学 | Effective dynamic commodity recommendation method and commodity recommendation system |
| CN102629360B (en) * | 2012-03-13 | 2016-04-20 | 浙江大学 | A kind of effective dynamic commodity recommend method and commercial product recommending system |
| CN102799623B (en) * | 2012-06-21 | 2015-07-22 | 杜小勇 | Information pushing method |
| CN102799623A (en) * | 2012-06-21 | 2012-11-28 | 杜小勇 | Information pushing method |
| CN103020855B (en) * | 2012-12-11 | 2016-02-10 | 北京京东世纪贸易有限公司 | The method and system of bad commodity is identified based on user's buying behavior |
| CN103020855A (en) * | 2012-12-11 | 2013-04-03 | 北京京东世纪贸易有限公司 | Bad commodity distinguishing method and system based on user purchasing behavior |
| CN103218407A (en) * | 2013-03-22 | 2013-07-24 | 南京信通科技有限责任公司 | Recommendation engine based on interest graph |
| CN103353865A (en) * | 2013-04-25 | 2013-10-16 | 杭州电子科技大学 | Barter electronic trading commodity recommendation method based on position |
| CN103353865B (en) * | 2013-04-25 | 2017-04-12 | 杭州电子科技大学 | Barter electronic trading commodity recommendation method based on position |
| CN103412865A (en) * | 2013-06-07 | 2013-11-27 | 广州唯品会信息科技有限公司 | Method and system for carrying out notification on website projects |
| CN103412865B (en) * | 2013-06-07 | 2016-08-31 | 广州唯品会信息科技有限公司 | The Notification Method of website item and system |
| CN103345503B (en) * | 2013-07-01 | 2016-04-13 | 杭州万事利丝绸科技有限公司 | A kind of silk product personalized recommendation method based on wavelet network |
| CN103345503A (en) * | 2013-07-01 | 2013-10-09 | 杭州万事利丝绸科技有限公司 | Silk product individualized recommendation method based on wavelet network |
| CN105096149B (en) * | 2014-05-09 | 2019-02-01 | 华为软件技术有限公司 | Business tine method for analyzing product association and device |
| CN105096149A (en) * | 2014-05-09 | 2015-11-25 | 华为软件技术有限公司 | Service content product correlation analysis method and apparatus |
| CN104899236A (en) * | 2014-11-13 | 2015-09-09 | 深圳市腾讯计算机系统有限公司 | Comment information display method, comment information display device and comment information display system |
| CN104899236B (en) * | 2014-11-13 | 2019-01-29 | 深圳市腾讯计算机系统有限公司 | A kind of comment information display methods, apparatus and system |
| CN104572880B (en) * | 2014-12-22 | 2018-03-02 | 中国科学院信息工程研究所 | The Parallel Implementation method and system of collaborative filtering based on user |
| CN104572880A (en) * | 2014-12-22 | 2015-04-29 | 中国科学院信息工程研究所 | Method and system for realizing concurrent cooperated filtering based on users |
| CN104809626A (en) * | 2015-03-17 | 2015-07-29 | 徐邑江 | Customized commodity recommending method based on user credit assessment |
| CN106547365A (en) * | 2015-09-17 | 2017-03-29 | 阿里巴巴集团控股有限公司 | The method and apparatus of commercial product recommending |
| CN105302880A (en) * | 2015-10-14 | 2016-02-03 | 合一网络技术(北京)有限公司 | Content correlation recommendation method and apparatus |
| CN105225135B (en) * | 2015-10-31 | 2021-05-18 | 广州华多网络科技有限公司 | Potential customer identification method and device |
| CN105225135A (en) * | 2015-10-31 | 2016-01-06 | 广州华多网络科技有限公司 | Potentiality customer recognition method and device |
| CN107644036A (en) * | 2016-07-21 | 2018-01-30 | 阿里巴巴集团控股有限公司 | A kind of method, apparatus and system of data object push |
| CN107644036B (en) * | 2016-07-21 | 2021-06-29 | 阿里巴巴集团控股有限公司 | Method, device and system for pushing data object |
| CN106777051A (en) * | 2016-12-09 | 2017-05-31 | 重庆邮电大学 | A kind of many feedback collaborative filtering recommending methods based on user's group |
| CN106651546B (en) * | 2017-01-03 | 2021-12-07 | 重庆邮电大学 | Electronic commerce information recommendation method oriented to smart community |
| CN106651546A (en) * | 2017-01-03 | 2017-05-10 | 重庆邮电大学 | Intelligent community oriented electronic commerce information recommendation method |
| CN109951724A (en) * | 2017-12-20 | 2019-06-28 | 阿里巴巴集团控股有限公司 | Recommended method, main broadcaster's recommended models training method and relevant device is broadcast live |
| CN109995837A (en) * | 2018-01-02 | 2019-07-09 | 中国移动通信有限公司研究院 | A kind of service package recommended method, device and server |
| CN109995837B (en) * | 2018-01-02 | 2021-11-19 | 中国移动通信有限公司研究院 | Service package recommendation method, device and server |
| CN109299349A (en) * | 2018-09-11 | 2019-02-01 | 广州视源电子科技股份有限公司 | Application recommendation method and device, equipment and computer-readable storage medium |
| CN109740914A (en) * | 2018-12-28 | 2019-05-10 | 武汉金融资产交易所有限公司 | A kind of method, storage medium, equipment and system that financial business is assessed, recommended |
| CN109857498A (en) * | 2019-01-09 | 2019-06-07 | 明基智能科技（上海）有限公司 | Intelligent content template recommender system and its method |
| CN111210253A (en) * | 2019-11-26 | 2020-05-29 | 恒大智慧科技有限公司 | Method, device and storage medium for matching consumer with sales consultant |
| CN111882400A (en) * | 2020-07-31 | 2020-11-03 | 平安国际融资租赁有限公司 | Behavior recognition analysis method and device, computer equipment and readable storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101853463A (en) | Collaborative filtering recommending method and system based on client characteristics | |
| Yang et al. | A survey of collaborative filtering based social recommender systems | |
| CN102629257B (en) | Commodity recommending method of e-commerce website based on keywords | |
| JP5596152B2 (en) | Information matching method and system on electronic commerce website | |
| CN102609533B (en) | Kernel method-based collaborative filtering recommendation system and method | |
| KR100961783B1 (en) | Apparatus and method for presenting personalized goods and vendors based on artificial intelligence, and recording medium thereof | |
| CN106600372A (en) | Commodity recommending method and system based on user behaviors | |
| CN102254028A (en) | Personalized commodity recommendation method and system integrating attributes and structural similarity | |
| CN103886487A (en) | Individualized recommendation method and system based on distributed B2B platform | |
| CN106600302A (en) | Hadoop-based commodity recommendation system | |
| CN106471491A (en) | A kind of collaborative filtering recommending method of time-varying | |
| CN104391849A (en) | Collaborative filtering recommendation method for integrating time contextual information | |
| CN102542490A (en) | Commodity recommendation method based on model matching | |
| CN101206751A (en) | Customer recommendation system based on data digging and method thereof | |
| CN102402757A (en) | Method and device for providing information, and method and device for determining comprehensive relevance | |
| KR101963817B1 (en) | Apparatus and method for generating prediction information based on a keyword search volume | |
| CN103559622A (en) | Characteristic-based collaborative filtering recommendation method | |
| US20220253874A1 (en) | Product evaluation system and method of use | |
| CN103309894B (en) | Based on search implementation method and the system of user property | |
| CN104820879A (en) | User behavior information analysis method and device thereof | |
| CN111310046A (en) | Object recommendation method and device | |
| CN112150227A (en) | Commodity recommendation method, system, device and medium | |
| CN109299426A (en) | A kind of recommended method and device of accurate top information | |
| CN102831526A (en) | Method and system for searching and sequencing commodities to carry out transaction on line | |
| CN110347923B (en) | Traceable fast fission type user portrait construction method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Open date: 20101006 |