CN106952130B - General article recommendation method based on collaborative filtering - Google Patents

General article recommendation method based on collaborative filtering Download PDF

Info

Publication number
CN106952130B
CN106952130B CN201710110168.1A CN201710110168A CN106952130B CN 106952130 B CN106952130 B CN 106952130B CN 201710110168 A CN201710110168 A CN 201710110168A CN 106952130 B CN106952130 B CN 106952130B
Authority
CN
China
Prior art keywords
recommendation
users
user
attribute
collaborative filtering
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
Application number
CN201710110168.1A
Other languages
Chinese (zh)
Other versions
CN106952130A (en
Inventor
周智恒
劳志辉
俞政
黄俊楚
代雨琨
李立军
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.)
South China University of Technology SCUT
Original Assignee
South China University of Technology SCUT
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 South China University of Technology SCUT filed Critical South China University of Technology SCUT
Priority to CN201710110168.1A priority Critical patent/CN106952130B/en
Publication of CN106952130A publication Critical patent/CN106952130A/en
Application granted granted Critical
Publication of CN106952130B publication Critical patent/CN106952130B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/06—Buying, selling or leasing transactions
    • G06Q30/0601—Electronic shopping [e-shopping]
    • G06Q30/0631—Item recommendations
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/953—Querying, e.g. by the use of web search engines
    • G06F16/9535—Search customisation based on user profiles and personalisation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/06—Buying, selling or leasing transactions
    • G06Q30/0601—Electronic shopping [e-shopping]
    • G06Q30/0623—Item investigation
    • G06Q30/0625—Directed, with specific intent or strategy

Abstract

The invention discloses a general article recommendation method based on collaborative filtering, which utilizes a collaborative filtering method based on users to calculate the similarity between the users through Euclidean distance to obtain a similar user set and further obtain candidate recommendation sets of different users; and then, calculating recommendation scores of the candidate recommendation sets by classifying and assigning the attributes of the initial interested articles of the user, thereby obtaining a recommendation result with strong applicability and robustness and helping the user to more conveniently obtain interested contents. Compared with the traditional recommendation method based on demographics and content, the method disclosed by the invention focuses more on individual difference among users, and develops the interest of the users for recommendation by analyzing the historical behavior data of the users; therefore, recommendation results are different from person to person, personal preferences of users are considered more, and different method parameters can be customized for different recommendation scenes.

Description

General article recommendation method based on collaborative filtering
Technical Field
The invention relates to the technical field of information recommendation, in particular to a general article recommendation method based on collaborative filtering.
Background
Now, the data explosion era has been entered, and with the development of Web2.0, the Web has become a platform for data sharing, so that it becomes more and more difficult for people to find the information they need in massive data.
In such a situation, search engines (Google, Bing, hundredths, etc.) are the best way to find target information quickly. When the user demands the user relatively clearly, the user can conveniently find the required information by searching the keywords by using the search engine. However, the search engine cannot fully satisfy the user's needs for information discovery because in many cases, the user does not actually specify his or her needs, or their needs are difficult to express with simple keywords. Or they need results that better meet their personal tastes and preferences, recommendation systems have emerged, corresponding to search engines, which are also commonly referred to as recommendation engines.
With the advent of recommendation engines, the way users obtain information has shifted from simple, well-targeted data searches to higher-level information discovery that better conforms to people's usage habits.
Today, with the development of recommendation technology, recommendation engines have been successful in both E-commerce (E-commerce, e.g., Amazon, current network) and some social-based sites (including music, movie and book sharing, e.g., bean, Mtime, etc.). This further illustrates that in the web2.0 environment, users need such information discovery mechanisms that are more intelligent and more aware of their needs, tastes, and preferences in the face of huge amounts of data. The collaborative filtering recommendation algorithm is the earliest and well-known recommendation algorithm. The main functions are prediction and recommendation. The algorithm discovers the preference of the user by mining the historical behavior data of the user, divides the user into groups based on different preferences and recommends commodities with similar tastes. Collaborative filtering recommendation algorithms are classified into two categories, namely user-based collaborative filtering algorithms (user-based collaborative filtering) and item-based collaborative filtering algorithms (item-based collaborative filtering). In brief, the following is: humans are classified as species and groups as groups.
Disclosure of Invention
The invention aims to solve the defects in the prior art and provides a general item recommendation method based on collaborative filtering.
The purpose of the invention can be achieved by adopting the following technical scheme:
a general item recommendation method based on collaborative filtering comprises the following steps:
s1, the service party performs initialization definition on the recommended articles and the attributes and characteristic values of the users, classifies the behaviors of the users, sets the conditions of an initial search engine, initiates a recommendation request to a recommendation engine APK after initialization setting and APK access of the recommendation engine are completed, and sends an initialized data set to the recommendation engine APK;
s2, a recommendation engine APK collects data meeting search conditions to form a recommendable candidate set A according to an initial search engine of a service party, a similar user set is obtained by similarity judgment of the recommendable candidate set A, and data screening is carried out according to a collaborative filtering thought based on users to obtain a recommendable result set B;
s3, dividing attribute feature values of the recommendable result set B according to attributes of the articles, obtaining the weight of each attribute feature value in the users of the recommendable result set B according to the proportion of each feature value in the interested article sets of the users, obtaining the attribute which is most sensitive to user perception according to each most significant feature set, and sequencing the interested article sets of the users according to different reference weights of the attribute which is most sensitive to perception to obtain a primary recommendation result list C;
and S4, sending the preliminary recommendation result list C to a service party, and rearranging the service party according to requirements to finally obtain a recommendation result list D.
Further, the specific process of step S2 is as follows:
s201, dividing the behaviors of the user into T1～TKTotally K classes, and respectively carrying out weight assignment w on the K classes of behaviors1～wkDividing the user behaviors into a positive dimension, a negative dimension, a high dimension, a middle dimension and a low dimension according to different user behaviors, wherein the value of an assignment vector w is w ═ 2, -1,0,1,2 and 3;
s202, acquiring a behavior operation accumulated value of the user on the article to obtain the preference degree H ═ Σ w of the user on the article; when H >3, then the user is considered to be interested in the item;
s203, utilizing Euclidean distance according to the preference degree H of different users to each article
Figure GDA0002568334310000031
Calculating to obtain the similarity between users:
Figure GDA0002568334310000032
and when the similarity sim (x, y) > k between the two users is determined by a service party, namely the two users are considered to be similar to obtain a similar user set, and a recommendable result set B of each user is obtained according to a collaborative filtering thought based on the users.
Further, the specific process of step S3 is as follows:
s301, according to the recommendable result set B, the attributes and the characteristic values of the articles are divided, and the attribute vector of the articles is set as:
Figure GDA0002568334310000033
attribute SiThe eigenvalue vector of (a) is:
Figure GDA0002568334310000034
s302, constructing an attribute feature matrix through the discrimination vectors of the attributes and the candidate recommendation sets;
for user A, the set of items of interest QAProperty S of a certain articleiCharacteristic value v ofkIn that
Figure GDA0002568334310000041
In the proportion of
Figure GDA0002568334310000042
Then in the candidate recommendation set QTIn the method, a characteristic value v of the article attribute is setkThe occupied weight is as follows:
Figure GDA0002568334310000043
in addition, when
Figure GDA0002568334310000044
(i, k take any value), then the attribute S is consideredxThe discrimination degree is strongest;and (3) taking the discrimination of each attribute:
Figure GDA0002568334310000045
therefore, when Q is givenAThen, a property region index vector is available:
Figure GDA0002568334310000046
to obtain QTThen, the attribute feature matrix of the article can be obtained:
Figure GDA0002568334310000047
s303, according to the attribute feature matrix and the attribute division vector, a candidate recommendation set Q can be obtainedTThe item recommendation score vector of (1):
Figure GDA0002568334310000048
s304, determining a preliminary recommendation result list C according to the obtained recommendation score vector to the sequence of the recommendable result set B.
Further, if the number of people who can recommend the candidate set a in the step S2 does not satisfy the minimum recommended number requirement, the business party is requested to expand the search condition.
Further, the attribute and the feature value of the user are defined in a mode of [ attribute-value ] key value pair.
Further, the initial attribute of the user is filled in by the user during registration, and the recommendation engine may perform feature division for the initial user according to the attribute value filled in by the user.
Compared with the prior art, the invention has the following advantages and effects:
compared with the traditional recommendation method based on demographics and content, the method disclosed by the invention draws on the collaborative filtering thought and the content-based recommendation thought, pays more attention to individual difference among users, and mines the interest of the users for recommendation by analyzing the historical behavior data of the users; therefore, recommendation results are different from person to person, personal preferences of users are considered more, and different method parameters can be customized for different recommendation scenes.
Drawings
FIG. 1 is a flow chart of a collaborative filtering based general item recommendation method disclosed in the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Examples
As shown in fig. 1, the embodiment discloses a general item recommendation method based on collaborative filtering, which designs a recommendation engine for a set of general items by using the idea of collaborative filtering and the idea of content-based recommendation. Wherein, general purpose articles include but are not limited to: books, music, movies, and different general articles are different in that the measurement standard of the preference and the attribute feature difference of the articles can be adjusted by the business party according to different recommended scenes.
The method for recommending articles specifically comprises the following steps:
s1, the service side carries out initialization definition on the attributes and the characteristic values of the recommended articles and the users, and the definition mode is an attribute-value key value pair, such as height-170 CM; classifying the behaviors of the users, setting conditions of an initial search engine, initiating a recommendation request to a recommendation engine APK after initialization setting and access of the recommendation engine APK are completed, and sending an initialized data set to the recommendation engine APK.
In specific application, the initial attribute of the user is filled in by the user during registration, and the recommendation engine can perform feature division for the initial user according to the attribute value filled in by the user.
S2, a recommendation engine APK collects data meeting search conditions to form a recommendable candidate set A according to an initial search engine of a service party, a similar user set is obtained by similarity judgment of the recommendable candidate set A, and data screening is carried out according to a collaborative filtering thought based on users to obtain a recommendable result set B; and if the number of people of the recommendable candidate set A does not meet the requirement of the minimum recommendable number of people, requesting the service party to expand the search condition.
The specific process of obtaining the recommendable result set B by data screening is as follows:
s201, dividing the behaviors of the user into T1～TKTotally K classes, and respectively carrying out weight assignment w on the K classes of behaviors1～wkDividing the user behaviors into a positive dimension, a negative dimension, a high dimension, a middle dimension and a low dimension according to different user behaviors, wherein the value of an assignment vector w is w ═ 2, -1,0,1,2 and 3;
in specific application, different behavior characteristics are defined according to online behavior data of a user.
S202, acquiring a behavior operation accumulated value of the user on the article to obtain the preference degree H ═ Σ w of the user on the article, and when H is greater than 3, considering that the user is interested in the article;
s203, utilizing Euclidean distance according to the preference degree H of different users to each article
Figure GDA0002568334310000061
Calculating to obtain the similarity between users:
Figure GDA0002568334310000062
and when the similarity sim (x, y) > k between the two users is determined by a service party, namely the two users are considered to be similar to obtain a similar user set, and a recommendable result set B of each user is obtained according to a collaborative filtering thought based on the users.
S3, the recommendable result set B is divided into attribute feature values according to the attributes of the articles, the weight of each attribute feature value in the users of the recommendable result set B is obtained according to the proportion of each feature value in the interested article sets of the users, the most sensitive attribute of user perception is obtained according to each most significant feature set, the interested article sets of the users are sorted according to different reference weights of the most sensitive attribute of perception, and a primary recommendation result list C is obtained.
The specific process of the step is as follows:
s301, according to the recommendable result set B, the attributes and the characteristic values of the articles are divided, and the attribute vector of the articles is set as:
Figure GDA0002568334310000071
attribute SiThe eigenvalue vector of (a) is:
Figure GDA0002568334310000072
s302, constructing an attribute feature matrix through the discrimination vectors of the attributes and the candidate recommendation sets;
for user A, the candidate recommendation set QTProperty S of a certain articleiCharacteristic value v ofkIn that
Figure GDA0002568334310000073
In the proportion of
Figure GDA0002568334310000074
Then in the candidate recommendation set QTIn the method, a characteristic value v of the article attribute is setkThe occupied weight is as follows:
Figure GDA0002568334310000075
in addition, when
Figure GDA0002568334310000076
(i, k are arbitrarily selected fromValue)
Then the attribute S is consideredxThe discrimination degree is strongest; and (3) taking the discrimination of each attribute:
Figure GDA0002568334310000077
therefore, when Q is givenAThen, a property region index vector is available:
Figure GDA0002568334310000081
to obtain QTThen, the attribute feature matrix of the article can be obtained:
Figure GDA0002568334310000082
s303, according to the attribute feature matrix and the attribute division vector, a candidate recommendation set Q can be obtainedTThe item recommendation score vector of (1):
Figure GDA0002568334310000083
s304, determining a preliminary recommendation result list C according to the obtained recommendation score vector to the sequence of the recommendable result set B.
And S4, sending the preliminary recommendation result list C to a service party, and carrying out appropriate rearrangement by the service party according to requirements to finally obtain a recommendation result list D, namely a final result.
In conclusion, the collaborative filtering thought and the content-based recommendation thought are used for reference, the individual difference among users is emphasized, and the interest of the users is mined for recommendation by analyzing the historical behavior data of the users; therefore, recommendation results are different from person to person, personal preferences of users are considered more, and different method parameters can be customized for different recommendation scenes.
The above embodiments are preferred embodiments of the present invention, but the present invention is not limited to the above embodiments, and any other changes, modifications, substitutions, combinations, and simplifications which do not depart from the spirit and principle of the present invention should be construed as equivalents thereof, and all such changes, modifications, substitutions, combinations, and simplifications are intended to be included in the scope of the present invention.

Claims (7)

1. A general item recommendation method based on collaborative filtering is characterized by comprising the following steps:
s1, the service party performs initialization definition on the recommended articles and the attributes and characteristic values of the users, classifies the behaviors of the users, sets the conditions of an initial search engine, initiates a recommendation request to a recommendation engine APK after initialization setting and APK access of the recommendation engine are completed, and sends an initialized data set to the recommendation engine APK;
s2, a recommendation engine APK collects data meeting search conditions to form a recommendable candidate set A according to an initial search engine of a service party, a similar user set is obtained by similarity judgment of the recommendable candidate set A, and data screening is carried out according to a collaborative filtering thought based on users to obtain a recommendable result set B;
s3, dividing attribute feature values of the recommendable result set B according to attributes of the articles, obtaining the weight of each attribute feature value of the users of the recommendable result set B according to the proportion of each feature value in the interested article sets of the users, obtaining the attribute which is most sensitive to user perception according to each most significant feature set, and sequencing the interested article sets of the users according to different reference weights of the attributes which are most sensitive to perception to obtain a preliminary recommendation result list C, wherein the specific process of the step S3 is as follows:
s301, according to the recommendable result set B, the attributes and the characteristic values of the articles are divided, and the attribute vector of the articles is set as:
Figure FDA0002615180490000011
attribute SiThe eigenvalue vector of (a) is:
Figure FDA0002615180490000012
s302, constructing an attribute feature matrix through the discrimination vectors of the attributes and the candidate recommendation sets;
for user A, user A's interest item set QAProperty S of a certain articleiCharacteristic value v ofkIn that
Figure FDA0002615180490000021
In the proportion of
Figure FDA0002615180490000022
Then in the candidate recommendation set QTIn the method, a characteristic value v of the article attribute is setkThe occupied weight is as follows:
Figure FDA0002615180490000023
in addition, when
Figure FDA0002615180490000024
if i and k take any value, the attribute S is consideredxThe discrimination degree is strongest; and (3) taking the discrimination of each attribute:
Figure FDA0002615180490000025
therefore, when Q is givenAThen, a property region index vector is available:
Figure FDA0002615180490000026
to obtain QTThen, the attribute feature matrix of the article can be obtained:
Figure FDA0002615180490000027
s303, distinguishing degree according to the attribute feature matrix and the attributeQuantity, candidate recommendation set Q can be obtainedTThe item recommendation score vector of (1):
Figure FDA0002615180490000028
s304, determining a preliminary recommendation result list C according to the obtained recommendation score vector to the sequence of the recommendable result set B;
and S4, sending the preliminary recommendation result list C to a service party, and rearranging the service party according to requirements to finally obtain a recommendation result list D.
2. The collaborative filtering-based general item recommendation method according to claim 1, wherein the specific process of step S2 is as follows:
s201, dividing the behaviors of the user into T1～TKTotally K classes, and respectively carrying out weight assignment w on the K classes of behaviors1～wkDividing the user behaviors into a positive dimension, a negative dimension, a high dimension, a middle dimension and a low dimension according to different user behaviors, wherein the value of an assignment vector w is w ═ 2, -1,0,1,2 and 3;
s202, acquiring a behavior operation accumulated value of the user on the article to obtain the preference degree H ═ Σ w of the user on the article;
s203, utilizing Euclidean distance according to the preference degree H of different users to each article
Figure FDA0002615180490000031
Calculating to obtain the similarity between users:
Figure FDA0002615180490000032
and when the similarity sim (x, y) > k between the two users is determined by a service party, namely the two users are considered to be similar to obtain a similar user set, and a recommendable result set B of each user is obtained according to a collaborative filtering thought based on the users.
3. The collaborative filtering based universal item recommendation method according to claim 1,
if the number of people who can recommend the candidate set a in the step S2 does not satisfy the minimum recommended number of people requirement, the business side is requested to expand the search condition.
4. The collaborative filtering-based universal item recommendation method according to claim 1, wherein the attributes and feature values of the user are defined in a manner of [ attribute-value ] key value pairs.
5. The collaborative filtering-based universal item recommendation method according to claim 1, wherein initial attributes of the users are filled in by the users at registration, and the recommendation engine performs feature classification for the initial users according to attribute values filled in by the users.
6. The collaborative filtering based universal item recommendation method according to claim 2,
when the user's preference for an item H >3, then the user is considered to be interested in the item.
7. The collaborative filtering based universal item recommendation method according to any one of claims 1 to 6,
the generic article comprises: books, music or movies, different general articles differ in that the measurement of the preference and the difference in the attribute characteristics of the articles can be adjusted by the business according to different recommended scenes.
CN201710110168.1A 2017-02-27 2017-02-27 General article recommendation method based on collaborative filtering Active CN106952130B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201710110168.1A CN106952130B (en) 2017-02-27 2017-02-27 General article recommendation method based on collaborative filtering

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201710110168.1A CN106952130B (en) 2017-02-27 2017-02-27 General article recommendation method based on collaborative filtering

Publications (2)

Publication Number Publication Date
CN106952130A CN106952130A (en) 2017-07-14
CN106952130B true CN106952130B (en) 2020-10-27

Family

ID=59468020

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201710110168.1A Active CN106952130B (en) 2017-02-27 2017-02-27 General article recommendation method based on collaborative filtering

Country Status (1)

Country Link
CN (1) CN106952130B (en)

Families Citing this family (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108038730A (en) * 2017-12-22 2018-05-15 联想(北京)有限公司 Product similarity determination methods, device and server cluster
CN110399185B (en) 2018-04-24 2022-05-06 华为技术有限公司 Method, terminal and server for adjusting intelligent recommendation
CN109214893A (en) * 2018-08-31 2019-01-15 深圳春沐源控股有限公司 Method of Commodity Recommendation, recommender system and computer installation
CN109345175B (en) * 2018-10-10 2021-08-03 江苏满运软件科技有限公司 Goods source pushing method, system, equipment and storage medium based on driver matching degree
CN111400532A (en) * 2019-01-03 2020-07-10 福建天泉教育科技有限公司 Photo recommendation method and terminal
CN112632405A (en) * 2020-12-31 2021-04-09 数字广东网络建设有限公司 Recommendation method, device, equipment and storage medium

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6041311A (en) * 1995-06-30 2000-03-21 Microsoft Corporation Method and apparatus for item recommendation using automated collaborative filtering
CN103023977A (en) * 2012-11-19 2013-04-03 华南理工大学 Recommendation system and method based on positional information
US8738467B2 (en) * 2006-03-16 2014-05-27 Microsoft Corporation Cluster-based scalable collaborative filtering
CN104915861A (en) * 2015-06-15 2015-09-16 浙江经贸职业技术学院 An electronic commerce recommendation method for a user group model constructed based on scores and labels
CN105095267A (en) * 2014-05-09 2015-11-25 阿里巴巴集团控股有限公司 User involving project recommendation method and apparatus
CN105976229A (en) * 2016-05-13 2016-09-28 云南大学 Collaborative filtering algorithm based on user and project mixing
CN106326483A (en) * 2016-08-31 2017-01-11 华南理工大学 Collaborative recommendation method with user context information aggregation
CN106339502A (en) * 2016-09-18 2017-01-18 电子科技大学 Modeling recommendation method based on user behavior data fragmentation cluster

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6041311A (en) * 1995-06-30 2000-03-21 Microsoft Corporation Method and apparatus for item recommendation using automated collaborative filtering
US8738467B2 (en) * 2006-03-16 2014-05-27 Microsoft Corporation Cluster-based scalable collaborative filtering
CN103023977A (en) * 2012-11-19 2013-04-03 华南理工大学 Recommendation system and method based on positional information
CN105095267A (en) * 2014-05-09 2015-11-25 阿里巴巴集团控股有限公司 User involving project recommendation method and apparatus
CN104915861A (en) * 2015-06-15 2015-09-16 浙江经贸职业技术学院 An electronic commerce recommendation method for a user group model constructed based on scores and labels
CN105976229A (en) * 2016-05-13 2016-09-28 云南大学 Collaborative filtering algorithm based on user and project mixing
CN106326483A (en) * 2016-08-31 2017-01-11 华南理工大学 Collaborative recommendation method with user context information aggregation
CN106339502A (en) * 2016-09-18 2017-01-18 电子科技大学 Modeling recommendation method based on user behavior data fragmentation cluster

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
信息推荐系统中的协同过滤技术研究;尹航;《万方学术论文数据库》;20151203;第41-56页 *

Also Published As

Publication number Publication date
CN106952130A (en) 2017-07-14

Similar Documents

Publication Publication Date Title
CN106952130B (en) General article recommendation method based on collaborative filtering
Reddy et al. Content-based movie recommendation system using genre correlation
US10402917B2 (en) Color-related social networking recommendations using affiliated colors
US11037222B1 (en) Dynamic recommendations personalized by historical data
Ning et al. A comprehensive survey of neighborhood-based recommendation methods
Cufoglu User profiling-a short review
TWI636416B (en) Method and system for multi-phase ranking for content personalization
WO2020048084A1 (en) Resource recommendation method and apparatus, computer device, and computer-readable storage medium
KR101098871B1 (en) APPARATUS AND METHOD FOR　MEASURING CONTENTS SIMILARITY BASED ON FEEDBACK INFORMATION OF RANKED USER and Computer Readable Recording Medium Storing Program thereof
US20120185481A1 (en) Method and Apparatus for Executing a Recommendation
CN108334632B (en) Entity recommendation method and device, computer equipment and computer-readable storage medium
KR101620748B1 (en) Item recommendation method and apparatus
CN106709076B (en) Social network recommendation device and method based on collaborative filtering
CN102073699A (en) Method, device and equipment for improving search result based on user behaviors
KR102108683B1 (en) Method for providing recommendation contents including non-interest contents
Feng et al. Improving group recommendations via detecting comprehensive correlative information
KR101708254B1 (en) Story-based recommendation system and method for movies by character-net and collaborative filtering
CN111310046A (en) Object recommendation method and device
Hong et al. Context-aware recommendation using role-based trust network
CN108647985B (en) Article recommendation method and device
CN110110206B (en) Method, device, computing equipment and storage medium for mining and recommending relationships among articles
KR101910424B1 (en) Method for movie ratings prediction using sentiment analysis of movie tags, recording medium and device for performing the method
US20150206220A1 (en) Recommendation Strategy Portfolios
WO2023206960A1 (en) Product recommendation method and apparatus based on content and collaborative filtering, and computer device
Usmani et al. A predictive approach for improving the sales of products in e-commerce

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant