CN103400286A - Recommendation system and method for user-behavior-based article characteristic marking - Google Patents

Recommendation system and method for user-behavior-based article characteristic marking Download PDF

Info

Publication number
CN103400286A
CN103400286A CN2013103335750A CN201310333575A CN103400286A CN 103400286 A CN103400286 A CN 103400286A CN 2013103335750 A CN2013103335750 A CN 2013103335750A CN 201310333575 A CN201310333575 A CN 201310333575A CN 103400286 A CN103400286 A CN 103400286A
Authority
CN
China
Prior art keywords
article
user behavior
user
correlation
degree
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.)
Granted
Application number
CN2013103335750A
Other languages
Chinese (zh)
Other versions
CN103400286B (en
Inventor
卢青峰
王冬杰
朱勇勇
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.)
Digital Trade Technology (Beijing) Co., Ltd.
Original Assignee
Century Light Technology Development (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 Century Light Technology Development (beijing) Co Ltd filed Critical Century Light Technology Development (beijing) Co Ltd
Priority to CN201310333575.0A priority Critical patent/CN103400286B/en
Publication of CN103400286A publication Critical patent/CN103400286A/en
Application granted granted Critical
Publication of CN103400286B publication Critical patent/CN103400286B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Landscapes

  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention relates to a recommendation system and a recommendation method for user-behavior-based article characteristic marking. The method comprises the following steps of constructing an implicit data sequence by using a searching behavior of a user and a behavior on an article after the searching behavior, extracting characteristics of the article, comparing basic data which is a calculation result of relevance between the characteristics and the article with a new behavior of the user, and recommending related products to the user after the related products are sequenced.

Description

A kind of commending system and method for carrying out the article characteristics mark based on user behavior
Technical field
The present invention relates to mobile internet technical field, be specifically related to a kind of online article commending system and implementation method thereof based on user behavior.
Background technology
Recommended technology be from a large amount of information, quick-searching may interested article to the user, and show a kind of technology of user, reduce the user retrieves in growing magnanimity information difficulty.
Current commending system, have two kinds of selections on the behavioral data that the user is produced is processed.A kind of is the demonstration feedback information of selecting the user, such as the user, the information such as the scoring of article, evaluation is used as the basic data of commending system.In some cases, the user can't provide the extra explicit feedback information such as comment, causes being used for the data of analyzing very sparse, makes commending system not reach the ideal effect of expection, is unfavorable for analyzing as basic data.Another is to adopt the mode that hidden data is analyzed, and obtains some users' behavior rule after hidden data is reorganized, and such as behaviors such as click, purchases, and does not need the user that extra feedback information is provided.For example the patent No. is to disclose a kind of terminal in 201210499328.3 Chinese patent application automatically to recommend method and the device of similar commodity, carry out information extraction by the commodity image to obtaining, searching out the commodity high with commodity similarity interested in database selects for the user, this patented claim can not carry out the search commercial articles database by the mode to obtain keyword, the mode of information extraction simultaneously is simple, and the commodity that therefore obtain from database can not meet user's request fully.
Summary of the invention
The object of the present invention is to provide a kind of method of how to recommend the basic data of article in online merchandise sales of processing: utilize computing machine, server and corresponding software basic data to be carried out the article vector of proper vector and the feature of cutting, data correlation, generation article, take out finally after the highest several commodity of similarity sort and recommend the user from merchandising database; Simultaneously, the present invention also provides the complete system of a cover coupling the method.
A first aspect of the present invention provides a kind of method of processing the basic data of recommendation:
1, collect user behavior data and be sent to terminal server from subscription client by network;
2, described user behavior data is sorted, and statistics time interval, this time interval is used for the cutting to this user behavior sequence;
The article that 3, will be arranged in same sequence carry out related with user behavior;
4, described user behavior is carried out feature extraction.
5, described article and described user behavior are generated its characteristic vector space, and calculate the score of each feature;
6,, to the characteristic vector space transposition of article, obtain the article vector space of feature;
7,, according to the article vector space of feature, calculate the degree of correlation of feature and article;
8, calculate the degree of correlation between article according to the characteristic vector space of article;
9, recommend the interested article of user according to degree of correlation size.
A second aspect of the present invention provides a kind of commending system that carries out the article characteristics mark based on user behavior, has three large data processing servers and three large memory modules, and data processing server has: preprocessing server, calculation server, recommendation server.Memory module has log store module, intermediate computations memory module, result store module.
Preferably, the user uses after certain search condition searches for, participated in the operation (operations such as click, purchase) of some article, if the probability that certain search condition and certain article occur simultaneously is larger, this search condition will be used as the user concealed attribute of these article.
Preferably, described implicit attribute obtains recommending item lists through the calculating of server and the storage of memory module in above-mentioned arbitrary scheme.
Preferably, described preprocessing server has 4 main modular in above-mentioned arbitrary scheme: user behavior data collection module, behavioral data molded tissue block, item associations behavior extraction module, article characteristics value computing module.
In above-mentioned arbitrary scheme preferably, the user behavior data collection module, be used for extracting user's behavioral data from described business diary memory module, and by the generation time of behavior, sequentially generate behavior sequence, every behavior record need to indicate its producer.
Preferably, described user behavior data collection module is sent to described intermediate computations memory module with every sequence and its producer in above-mentioned arbitrary scheme, the concrete user behavior memory module that is sent under this module.
Preferably, described intermediate computations memory module also has article characteristics vector memory module, article characteristics data memory module in above-mentioned arbitrary scheme.
Preferably, described user behavior memory module is used for the necessary user behavior data of storage in above-mentioned arbitrary scheme.
Preferably, the user behavior data that described user behavior memory module obtains this module is sent to described user behavior data molded tissue block in above-mentioned arbitrary scheme.
Preferably, the function that described behavioral data molded tissue block has comprises in above-mentioned arbitrary scheme: in the statistics behavior cutting time interval, the time interval that obtains according to statistics is cut into behavior segment with the user behavior sequence.
In above-mentioned arbitrary scheme preferably, the fragment that described behavioral data molded tissue block will obtain is carried out related to article and user behavior data by described item associations behavior extraction module, and with user behavior data by after regular cutting as the feature of article.
Preferably, described item associations behavior extraction module is sent to described article characteristics vector memory module with article characteristics and stores in above-mentioned arbitrary scheme.
In above-mentioned arbitrary scheme preferably, the information extraction from described article characteristics vector memory module of described article characteristics value computing module.
Preferably, described article characteristics value computing module, obtain the score of each feature of each article after being used for the feature frequency of occurrences is added up and calculated, and the characteristic storage that these implicit feedback generate is become the proper vector of article in above-mentioned arbitrary scheme.
Preferably, the proper vector of described article is stored in described article characteristics data memory module in above-mentioned arbitrary scheme.
Preferably, described calculation server has relatedness computation module between relatedness computation module, article and the article of feature and article in above-mentioned arbitrary scheme.
Preferably, the information in the described article characteristics data memory module of described calculation server reception is called different computing modules by situation and is calculated in above-mentioned arbitrary scheme.
Preferably, the relatedness computation module of described feature and article, adopt the conditional probability of Bayesian formula calculated characteristics corresponding to article, according to this feature, looks for its corresponding article, by the probability sizes values, sorts in above-mentioned arbitrary scheme.
Preferably, relatedness computation module between described article and article, adopt and revise cosine degree of correlation formula, in the degree of correlation of calculating on user's implicit features dimension between article in above-mentioned arbitrary scheme.
Preferably, the sequence that the relatedness computation module of described feature and article obtains stores described result store module in above-mentioned arbitrary scheme, concrete feature and the article degree of correlation memory module that is stored under this module.
Preferably, described result store module also has degree of correlation memory module between article and article in above-mentioned arbitrary scheme, is used for storing coming from the article degree of correlation result that between described article and article, the relatedness computation module obtains.
Preferably, described recommendation server is used for receiving user's request in above-mentioned arbitrary scheme, and returns to the product of recommending the user after request is resolved.
Preferably, described recommendation server has the recommendation of processing two kinds of request modes, search recommending module and relative article recommending module in above-mentioned arbitrary scheme.
Preferably, when described search recommending module is user's input search condition, be its recommendation product relevant to search condition in above-mentioned arbitrary scheme.
Preferably, described relative article recommending module is when the user operates a certain article in above-mentioned arbitrary scheme, for it recommends the relevant article of article therewith.
Description of drawings
Fig. 1 is the structural representation according to a preferred embodiment of the commending system that carries out article characteristics mark based on user behavior of the present invention.
Embodiment
Below in conjunction with the drawings and specific embodiments, technical scheme of the present invention is described in detail.
The object of the present invention is to provide a kind of method of processing the basic data of recommendation, and be based upon the system on the method.Build the hidden data sequence by the behavior that is added on article after user's search behavior and search behavior, and use search condition to redefine the feature of article, the degree of correlation between the degree of correlation between feature and article and article and article is calculated, and two kinds of results will calculating are stored respectively the basic data of using when conduct is recommended.When the user searches for, user's search condition is decomposed, the word that decomposition is obtained mates with the feature that calculates and the data of the article degree of correlation, obtains the list of relative article, according to recommending the user after relevancy ranking; When the user operates a certain article, get according to the article of current operation in the degree of correlation data between the article that calculate and inquire about, obtain the list of relative article, recommend the user after sequence.
Data processing method provided by the invention is as follows:
1, collect user behavior data and be sent to terminal server from subscription client by network;
2, by program, described user behavior data is sorted on processing server, and statistics time interval, this time interval is used for the cutting of user behavior sequence;
The article that 3, will be arranged in same sequence carry out related with user behavior;
4, described user behavior is carried out feature extraction.
5, described article and described user behavior are generated its characteristic vector space, and calculate the score of each feature;
6,, to the characteristic vector space transposition of article, obtain the article vector space of feature;
7,, according to the article vector space of feature, calculate the degree of correlation of feature and article;
8, calculate the degree of correlation between article according to the characteristic vector space of article;
9, recommend the interested article of user according to degree of correlation size.
Shown in said method, a kind of specific embodiment of this invention is as follows:
A kind of commending system that carries out the article characteristics mark based on user behavior, this system comprises 3 large data processing servers and 3 large memory modules.Data processing server comprises preprocessing server, calculation server, recommendation server.Memory module comprises log store module, intermediate computations memory module, result store module.Described preprocessing server comprises 4 main modular: user behavior data collection module, behavioral data molded tissue block, item associations behavior extraction module, article characteristics value computing module.
As shown in Figure 1, described user behavior data collection module, be used for extracting user's behavioral data from described business diary memory module, and by the generation time of behavior, sequentially generate behavior sequence, and every behavior record need to indicate its producer.The behavioral data of collecting comprises: the article of user search condition and user's operation, the commodity that the commodity of browsing such as the user or user buy.
The user behavior data that described user behavior data collection module will obtain is stored in described user behavior memory module, then described user behavior memory module arrives the behavioral data molded tissue block with information conveyance, described behavioral data molded tissue block has following function: the statistics behavior cutting time interval, the time interval that obtains according to statistics is cut into behavior segment with the user behavior sequence, describedly based on user behavior, carries out the commending system of article characteristics mark and record that method is set in behavior segment is that certain causal relation is arranged.
Described behavioral data molded tissue block is after carrying out cutting to the user behavior sequence, fragment after cutting is carried out related extracting section, this is operated in described item associations behavior extraction module and carries out, described item associations behavior extraction module, be used for article and user behavior data are carried out related, and with user behavior data by regular cutting afterwards as the feature of article.Such as user search keyword " mp3 bag ", then browsed again item1 in same behavior fragment, carry out related with searching key word mp3 bag item1 so, mp3 and bag are as the user concealed feature that article item1 is given, no matter and whether comprise these keywords in the description content of article providers to item1.
Described item associations behavior extraction module with keyword and article carry out related after, be stored in described article characteristics vector memory module, then described article characteristics value computing module calls described article characteristics vector memory module canned data for the score that obtains each feature of each article after the feature frequency of occurrences is added up and calculated, and the characteristic storage that these implicit feedback generate is become the proper vector of article, I<F1:S1, F2:S2 ... 〉, wherein I is article, F1 is the feature of I, and S1 is the score of F1.For example two of item1 key feature mp3 and bag, its vectorial score is described below:
Item1<mp3:0.3, bag:0.6 〉, item1 has two features, wherein feature " mp3 " must be divided into 0.3, wherein feature " bag " must be divided into 0.6.
The vector that described article characteristics value computing module will calculate is stored in described article characteristics data memory module, then described calculation server extracts data from described article characteristics data memory module, and this calculation server has relatedness computation module between the relatedness computation module of feature and article and article and article.
The relatedness computation module of described feature and article is used for calculating the degree of correlation of special this and article, adopts the conditional probability of Bayesian formula calculated characteristics corresponding to article:
P(Item|F)=P(Item)P(F|Item)/P(F)
When being used for doing the recommendation article,, if there is this feature, so according to this feature, look for its corresponding article, sort by the probability sizes values.When as the user, inputting keyword mp3, the relatedness computation module of described feature and article will be to article item1, item2 ... carry out probability calculation, draw P (item1) and P (item2) ... then comparative feature judges with this whether article appear in recommendation list corresponding to the conditional probability size of article.
Relatedness computation module between described article and article, adopt and revise cosine degree of correlation formula, in the degree of correlation of calculating on user's implicit features dimension between article.Revising cosine degree of correlation formula adopts:
Figure 2013103335750100002DEST_PATH_IMAGE001
This formula be used for to calculate the degree of correlation between i article and j article, is designated as sim(i, j) wherein
Figure 599560DEST_PATH_IMAGE002
Represent i article feature f score value,
Figure 2013103335750100002DEST_PATH_IMAGE003
Score mean value for feature f.F is characteristic set
Two kinds of results that described calculation server calculates will be kept at respectively in described result store module, and this result store module comprises that two memory modules are to store dissimilar result data: degree of correlation memory module between feature and article degree of correlation memory module, article and article.Wherein, feature and article degree of correlation memory module are used for the sequence that storage feature and article relatedness computation module obtain, and its storage format is: Feature<Item1:score1, Item2:score2... 〉; Wherein, Feature is feature, and the article of this feature association comprise Item1, Item2 ... and each related article has corresponding degree of association score, namely similarity.Between article and article, degree of correlation memory module is used for the sequence as a result that between stores and article, the relatedness computation module obtains, its storage format is: Item<Item1:score1, Item2:score2... 〉, wherein Item is the article of request, Item1, Item2 ... for the article that are associated with article Item, and provide score score.
Described recommendation server is used for carrying out Products Show to the user, and at first this server receives user's request, then keyword or the article of user's input is resolved, and namely user's request analysis, return to the product of recommending the user finally.Described user asks the concrete steps of explaining to carry out cutting for the keyword of the user being asked time input, and segmentation rules is consistent with the feature slit mode in item associations behavior extraction module: with this keyword or sentence segmentation, be word or phrase.
Described recommendation server comprises the two kinds of request modes of processing, a kind of while being user's input search condition such as keyword or one section statement of describing product, be that it recommends the product relevant to search condition; Another is when the user operates a certain article, for it recommends the relevant article of article therewith.
Need to prove; carry out the commending system of article characteristics mark and method based on user behavior and comprise any one and combination in any thereof in above-described embodiment according to of the present invention; but embodiment recited above is described the preferred embodiment of the present invention; not the scope of the invention is limited; design under spiritual prerequisite not breaking away from the present invention; various distortion and improvement that the common engineering technical personnel in this area make technical scheme of the present invention, all should fall in the definite protection domain of claims of the present invention.

Claims (10)

1. online method of recommending similar commodity, the method is sent to user behavior on terminal server by network, described terminal server is processed this user behavior and is obtained data cell, then return to again client according to the similar article of data cell retrieval from database, it is characterized in that: described processing user behavior comprises that terminal is logical user behavior is carried out feature extraction with related, and the generating feature vector space, the similar article of described retrieval comprise that the similar article of taking-up are as recommending article and these recommendation article being sorted from database.
2. the method for the similar commodity of online recommendation according to claim 1 is characterized in that: describedly user behavior is carried out feature extraction with related implementation, comprise the steps:
A, collection user behavior data;
B, user behavior is sorted, and statistics time interval, this time interval is used for the cutting of user behavior sequence;
C, described article are related with described user behavior;
D, described user behavior is carried out described feature extraction.
3. the method for the similar commodity of online recommendation according to claim 2 is characterized in that: describedly user behavior is carried out described feature extraction comprise the degree of correlation of calculating article and article.
4. the method for the similar commodity of online recommendation according to claim 3 is characterized in that: describedly user behavior is carried out described feature extraction also comprise the degree of correlation of calculated characteristics and article.
5. the method for the similar commodity of online recommendation according to claim 4, it is characterized in that: the concrete steps of the degree of correlation between described calculating article and article are:
A, described search condition is carried out feature extraction after, article are generated its characteristic vector space, be called the characteristic vector space of article, and calculate the score of each feature;
B, according to the characteristic vector space of article, calculate the degree of correlation between article and article.
6. the method for the similar commodity of online recommendation according to claim 5, it is characterized in that: the concrete steps of the degree of correlation of described calculated characteristics and article are:
A, the proper vector of described article is carried out the space transposition, obtain the article vector space of feature;
B, according to the article vector space of described feature, calculate, obtain the degree of correlation of feature and article.
7. the method for the similar commodity of online recommendation according to claim 6 is characterized in that: the relatedness computation method of described feature and article adopts Bayesian formula to calculate.
8. the method for the similar commodity of online recommendation according to claim 5 is characterized in that: between described article and article, the relatedness computation method adopts and revises cosine degree of correlation formula and calculate.
9. the method for the similar commodity of online recommendation according to claim 6, it is characterized in that: described to recommending the method that article sort to comprise: according to the degree of correlation between the degree of correlation of described feature and article and article and article, take out similar commodity from database, sort according to degree of correlation size, recommend the user.
10. online system of recommending similar commodity, by user behavior is obtained, then retrieve in merchandising database, obtain like product, and it is recommended the user, it is characterized in that: described article commending system has data processing server and data memory module, and by server, user behavior is carried out cutting, and the information of getting from database is sorted.
CN201310333575.0A 2013-08-02 2013-08-02 A kind of commending system and method carrying out article characteristics mark based on user behavior Active CN103400286B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201310333575.0A CN103400286B (en) 2013-08-02 2013-08-02 A kind of commending system and method carrying out article characteristics mark based on user behavior

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201310333575.0A CN103400286B (en) 2013-08-02 2013-08-02 A kind of commending system and method carrying out article characteristics mark based on user behavior

Publications (2)

Publication Number Publication Date
CN103400286A true CN103400286A (en) 2013-11-20
CN103400286B CN103400286B (en) 2016-12-28

Family

ID=49563901

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201310333575.0A Active CN103400286B (en) 2013-08-02 2013-08-02 A kind of commending system and method carrying out article characteristics mark based on user behavior

Country Status (1)

Country Link
CN (1) CN103400286B (en)

Cited By (18)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103700011A (en) * 2014-01-13 2014-04-02 重庆大学 Feature extraction method and device
CN103942712A (en) * 2014-05-09 2014-07-23 北京联时空网络通信设备有限公司 Product similarity based e-commerce recommendation system and method thereof
CN103995831A (en) * 2014-04-18 2014-08-20 新浪网技术(中国)有限公司 Object processing method, system and device based on similarity among objects
CN104063589A (en) * 2014-06-16 2014-09-24 百度移信网络技术(北京)有限公司 Recommendation method and system
CN104156450A (en) * 2014-08-15 2014-11-19 同济大学 Item information recommending method based on user network data
CN104750760A (en) * 2013-12-31 2015-07-01 中国移动通信集团上海有限公司 Application software recommending method and device
CN106649714A (en) * 2016-12-21 2017-05-10 重庆邮电大学 topN recommendation system and method for data non-uniformity and data sparsity
CN106708821A (en) * 2015-07-21 2017-05-24 广州市本真网络科技有限公司 User personalized shopping behavior-based commodity recommendation method
CN106959966A (en) * 2016-01-12 2017-07-18 腾讯科技(深圳)有限公司 A kind of information recommendation method and system
CN108399550A (en) * 2017-02-07 2018-08-14 北京京东尚科信息技术有限公司 A kind of tenant group method
CN109214848A (en) * 2017-07-06 2019-01-15 云义科技股份有限公司 Method and system for analyzing influence similarity of virtual commodities on recommendation system
CN109885644A (en) * 2019-04-08 2019-06-14 浙江大学城市学院 A kind of importance appraisal procedure for Internet of Things Item Information searching order
CN110782287A (en) * 2019-10-25 2020-02-11 北京沃东天骏信息技术有限公司 Entity similarity calculation method and device, article recommendation system, medium and equipment
CN111027895A (en) * 2019-05-16 2020-04-17 珠海随变科技有限公司 Stock prediction and behavior data collection method, apparatus, device and medium for commodity
CN111461841A (en) * 2020-04-07 2020-07-28 腾讯云计算(北京)有限责任公司 Article recommendation method, device, server and storage medium
CN111488385A (en) * 2020-04-07 2020-08-04 腾讯科技(深圳)有限公司 Data processing method and device based on artificial intelligence and computer equipment
CN111767458A (en) * 2019-09-11 2020-10-13 北京京东尚科信息技术有限公司 Information pushing method, device, system and storage medium
CN112861684A (en) * 2021-01-29 2021-05-28 北京电解智科技有限公司 Article display method and device

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101206752A (en) * 2007-12-25 2008-06-25 北京科文书业信息技术有限公司 Electric commerce website related products recommendation system and method
US20090248494A1 (en) * 2008-04-01 2009-10-01 Certona Corporation System and method for collecting and targeting visitor behavior
CN102509233A (en) * 2011-11-29 2012-06-20 汕头大学 User online action information-based recommendation method

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101206752A (en) * 2007-12-25 2008-06-25 北京科文书业信息技术有限公司 Electric commerce website related products recommendation system and method
US20090248494A1 (en) * 2008-04-01 2009-10-01 Certona Corporation System and method for collecting and targeting visitor behavior
CN102509233A (en) * 2011-11-29 2012-06-20 汕头大学 User online action information-based recommendation method

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
A.ELBADRAWY 等: "User-Specific Feature-based Similarity Models for Top-n Recommendation of New Items", 《ACM TRANSACTIONS ON INTELLIGENT SYSTEMS AND TECHNOLOGY》 *

Cited By (28)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104750760A (en) * 2013-12-31 2015-07-01 中国移动通信集团上海有限公司 Application software recommending method and device
CN104750760B (en) * 2013-12-31 2018-11-23 中国移动通信集团上海有限公司 A kind of implementation method and device for recommending application software
CN103700011A (en) * 2014-01-13 2014-04-02 重庆大学 Feature extraction method and device
CN103700011B (en) * 2014-01-13 2016-11-23 重庆大学 A kind of Feature Extraction Method and device
CN103995831B (en) * 2014-04-18 2017-04-12 新浪网技术(中国)有限公司 Object processing method, system and device based on similarity among objects
CN103995831A (en) * 2014-04-18 2014-08-20 新浪网技术(中国)有限公司 Object processing method, system and device based on similarity among objects
CN103942712A (en) * 2014-05-09 2014-07-23 北京联时空网络通信设备有限公司 Product similarity based e-commerce recommendation system and method thereof
CN104063589A (en) * 2014-06-16 2014-09-24 百度移信网络技术(北京)有限公司 Recommendation method and system
CN104063589B (en) * 2014-06-16 2018-01-16 百度移信网络技术(北京)有限公司 A kind of recommendation method and system
CN104156450B (en) * 2014-08-15 2017-11-07 同济大学 A kind of Item Information based on user network data recommends method
CN104156450A (en) * 2014-08-15 2014-11-19 同济大学 Item information recommending method based on user network data
CN106708821A (en) * 2015-07-21 2017-05-24 广州市本真网络科技有限公司 User personalized shopping behavior-based commodity recommendation method
CN106959966A (en) * 2016-01-12 2017-07-18 腾讯科技(深圳)有限公司 A kind of information recommendation method and system
CN106649714A (en) * 2016-12-21 2017-05-10 重庆邮电大学 topN recommendation system and method for data non-uniformity and data sparsity
CN108399550B (en) * 2017-02-07 2021-05-25 北京京东尚科信息技术有限公司 User grouping method
CN108399550A (en) * 2017-02-07 2018-08-14 北京京东尚科信息技术有限公司 A kind of tenant group method
CN109214848A (en) * 2017-07-06 2019-01-15 云义科技股份有限公司 Method and system for analyzing influence similarity of virtual commodities on recommendation system
CN109214848B (en) * 2017-07-06 2020-10-27 云义科技股份有限公司 Method and system for analyzing influence similarity of virtual commodities on recommendation system
CN109885644B (en) * 2019-04-08 2021-04-06 浙江大学城市学院 Importance evaluation method for searching and sorting of Internet of things item information
CN109885644A (en) * 2019-04-08 2019-06-14 浙江大学城市学院 A kind of importance appraisal procedure for Internet of Things Item Information searching order
CN111027895A (en) * 2019-05-16 2020-04-17 珠海随变科技有限公司 Stock prediction and behavior data collection method, apparatus, device and medium for commodity
CN111767458A (en) * 2019-09-11 2020-10-13 北京京东尚科信息技术有限公司 Information pushing method, device, system and storage medium
CN110782287A (en) * 2019-10-25 2020-02-11 北京沃东天骏信息技术有限公司 Entity similarity calculation method and device, article recommendation system, medium and equipment
CN111461841A (en) * 2020-04-07 2020-07-28 腾讯云计算(北京)有限责任公司 Article recommendation method, device, server and storage medium
CN111488385A (en) * 2020-04-07 2020-08-04 腾讯科技(深圳)有限公司 Data processing method and device based on artificial intelligence and computer equipment
CN111461841B (en) * 2020-04-07 2023-04-07 腾讯云计算(北京)有限责任公司 Article recommendation method, device, server and storage medium
CN111488385B (en) * 2020-04-07 2023-08-15 腾讯科技(深圳)有限公司 Data processing method and device based on artificial intelligence and computer equipment
CN112861684A (en) * 2021-01-29 2021-05-28 北京电解智科技有限公司 Article display method and device

Also Published As

Publication number Publication date
CN103400286B (en) 2016-12-28

Similar Documents

Publication Publication Date Title
CN103400286A (en) Recommendation system and method for user-behavior-based article characteristic marking
US9767166B2 (en) System and method for predicting user behaviors based on phrase connections
CN107729336B (en) Data processing method, device and system
EP2823410B1 (en) Entity augmentation service from latent relational data
US10268670B2 (en) System and method detecting hidden connections among phrases
Yakout et al. Infogather: entity augmentation and attribute discovery by holistic matching with web tables
CN106959971B (en) User behavior data processing method and device
JP5721818B2 (en) Use of model information group in search
US8983971B2 (en) Method, apparatus, and system for mobile search
US9934293B2 (en) Generating search results
CN108363821A (en) A kind of information-pushing method, device, terminal device and storage medium
CN104933100A (en) Keyword recommendation method and device
CN103838756A (en) Method and device for determining pushed information
CN104885081A (en) Search system and corresponding method
CN103279513A (en) Method for generating content label and method and device for providing multi-media content information
CN108763321A (en) A kind of related entities recommendation method based on extensive related entities network
Huynh et al. Scientific publication recommendations based on collaborative citation networks
Liu et al. Recommending blog articles based on popular event trend analysis
CN108205572A (en) A kind of searching method, device and equipment
CN107193883B (en) Data processing method and system
Diaby et al. Taxonomy-based job recommender systems on Facebook and LinkedIn profiles
Xu et al. Web content mining
Kantorski et al. Automatic filling of hidden web forms: A survey
Rajkumar et al. Users’ click and bookmark based personalization using modified agglomerative clustering for web search engine
TWI735516B (en) Method and device for processing user behavior data

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C14 Grant of patent or utility model
GR01 Patent grant
TR01 Transfer of patent right

Effective date of registration: 20180731

Address after: 100080 1-701-1 seven, level 28, Cheng Fu Road, Haidian District, Beijing.

Patentee after: Digital Trade Technology (Beijing) Co., Ltd.

Address before: 100083 Beijing Haidian District Huayuan Road 3-2, 6 floor of Dimon building.

Patentee before: Century Light Technology Development (Beijing) Co., Ltd.

TR01 Transfer of patent right