CN109101667A - A kind of personalized recommendation method based on explicit trust and implicit trust - Google Patents
A kind of personalized recommendation method based on explicit trust and implicit trust Download PDFInfo
- Publication number
- CN109101667A CN109101667A CN201811152650.2A CN201811152650A CN109101667A CN 109101667 A CN109101667 A CN 109101667A CN 201811152650 A CN201811152650 A CN 201811152650A CN 109101667 A CN109101667 A CN 109101667A
- Authority
- CN
- China
- Prior art keywords
- user
- trust
- project
- degree
- prediction
- 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
Links
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The present invention relates to field of computer technology, more particularly to a kind of based on the explicit personalized recommendation method trusted with implicit trust.The present invention first according to users to trust rating matrix, calculates the trust similarity between user, screens and trusts neighbours' collection recently;Then the explicit degree of belief trusted between neighbours concentration user recently is calculated;Calculate the prediction degree of belief between user;It predicts to score according to explicit trusting relationship;According to user-project rating matrix, the implicit trust degree between user is calculated, user concealed trust matrix is obtained;It is scored according to implicit trust Relationship Prediction;The two is merged, prediction target user is to the score value of project, and descending arranges, and prediction is scored highest K project recommendation to target user, generation recommendation list.Show that inventive can improve the accuracy rate and precision of recommendation, considerably reduce the error of recommendation better than current recommended method finally, comparing by experiment, is able to solve the problem for passing Deta sparseness and cold start-up.
Description
Technical field
The present invention relates to field of computer technology, more particularly to a kind of based on the explicit personalization trusted with implicit trust
Recommended method.
Background technique
With the development of Web3.0, Internet technology is constantly updated, so that the data explosion formula in network increases, is occurred
The problem of information overload, then to find that valuable information is just more difficult from mass data, the development of recommended technology
The preference tendency that user according to the history score information of user, can be excavated, to carry out personalized recommendation for user, still
There are the problems of Deta sparseness and cold start-up for traditional personalized recommendation method, and the accuracy of recommendation results is low, mistake
Rate is high.
Current personalized recommendation method is recommended performance poor, was drawn on this basis later there is recommending accuracy not high
The trusting relationship between user is entered, has improved the performance of recommender system to a certain extent, but has only accounted for straight between user
Connect trusting relationship, do not account for the explicit or implicit trusting relationship between user, thus how to make full use of explicit between user or
Implicit trust relationship improves the accuracy of recommendation, reduces and recommends error rate, even more there is an urgent need to one of hot spots of research.
Summary of the invention
The purpose of the present invention is to provide a kind of based on the explicit personalized recommendation method trusted with implicit trust.
The purpose of the present invention is what is realized by following approach: a kind of to be pushed away based on the explicit personalization trusted with implicit trust
Recommend method, it is assumed that existing subscriber-project rating matrix and users to trust rating matrix include the following steps:
Step 1: according to users to trust rating matrix, the trust similarity between user is calculated using formula (1),
And arrange in descending order, it chooses and trusts neighbours' collection recently with the highest top n user composition of user c trust similitude;
Wherein,Indicate the trust similarity between user c and user d,It is user c and user d
The user's number trusted jointly,It is the sum of user in users to trust rating matrix;
Step 2: trusting neighbours' collection recently according to obtained in step 1, calculated using formula (2) and trust neighbours concentration user recently
Between explicit degree of belief;
Wherein,Indicate user c to the explicit degree of belief of user d,Indicate the in-degree of user d,Table
Show maximum in-degree in all users of user c trust;
Step 3: the explicit degree of belief between the user according to obtained in step 2 calculates the prediction between user using formula (3)
Degree of belief;
Wherein,It is the prediction degree of belief between user c and user d,Be with user c similitude most
High top n trusts neighbours' collection recently,It is the trust similarity between user c and user q,
Indicate user q to the explicit degree of belief of user d;
Step 4: the prediction degree of belief between user generated according to step 3 calculates target user to project using formula (4)
Prediction scoring;
Wherein,It is in explicit trusting relationship, user c scores to the prediction of project i,It is user c and user d
Between prediction degree of belief,It is score value of the user d to project i,Indicate the number of users in users to trust rating matrix
Amount;
Step 5: according to user-project rating matrix, calculating the implicit trust degree between user using formula (5), obtain user
Implicit trust matrix,
Wherein,It is the implicit trust degree between user c and user d,It is score value of the user c to project i,
Indicate that user c comments excessive all items,It indicates that user d had project i and is selected as 1, do not selected to be 0,
Indicate that user d comments the number of excessive project;
Step 6: the user concealed trust matrix obtained according to step 5 calculates target user to the pre- of project using formula (6)
Assessment point;
Wherein,It is in implicit trust relationship, user c scores to the prediction of project i,It is user c and user d
Between implicit trust degree,It is score value of the user d to project i,Indicate the number of users in user-project rating matrix
Amount;
Step 7: for root according to step 4 and step 6 as a result, being merged using formula (7), prediction target user comments project
Score value, and descending arrangement is carried out, prediction is scored highest K project recommendation to target user, generates recommendation list;
Wherein,It is prediction score value of the target user c to project i,It is fusion factor, its value range is (0,1),It is in explicit trusting relationship, user c scores to the prediction of project i,It is in implicit trust relationship, user c is to project i's
Prediction scoring.
The present invention is a kind of based on the explicit personalized recommendation method trusted with implicit trust, compared with prior art, this hair
Bright elder generation calculates the trust similarity between user according to users to trust rating matrix, screens and trusts neighbours' collection recently;Then it calculates most
The nearly explicit degree of belief trusted between neighbours concentration user;Calculate the prediction degree of belief between user;It is pre- according to explicit trusting relationship
Assessment point;According to user-project rating matrix, the implicit trust degree between user is calculated, user concealed trust matrix is obtained;Root
It scores according to implicit trust Relationship Prediction;The two to be merged, prediction target user is to the score value of project, and descending arranges,
Prediction is scored highest K project recommendation to target user, generates recommendation list.Show this hair finally, comparing by experiment
The a kind of of bright proposition is trusted with the personalized recommendation method performance of implicit trust based on explicit better than current recommended method, is improved
The accuracy rate and precision recommended, considerably reduce the error of recommendation, are able to solve and pass Deta sparseness and cold start-up
Problem.
Detailed description of the invention
The present invention is described in further detail below in conjunction with the accompanying drawings:
Fig. 1 is flow diagram of the present invention.
Specific embodiment
As shown in Figure 1, the present invention is a kind of based on the explicit personalized recommendation method trusted with implicit trust, it is assumed that useful
Family-project rating matrix and users to trust rating matrix, include the following steps:
Step 1: according to users to trust rating matrix, the trust similarity between user being calculated using formula (1), and press
Descending arrangement is chosen and trusts neighbours' collection recently with the highest top n user composition of user c trust similitude;
Wherein,Indicate the trust similarity between user c and user d,It is user c and user d
The user's number trusted jointly,It is the sum of user in users to trust rating matrix;
Step 2: trusting neighbours' collection recently according to obtained in step 1, calculated using formula (2) and trust neighbours concentration user recently
Between explicit degree of belief;
Wherein,Indicate user c to the explicit degree of belief of user d,Indicate the in-degree of user d,Table
Show maximum in-degree in all users of user c trust;
Step 3: the explicit degree of belief between the user according to obtained in step 2 calculates the prediction between user using formula (3)
Degree of belief;
Wherein,It is the prediction degree of belief between user c and user d,Be with user c similitude most
High top n trusts neighbours' collection recently,It is the trust similarity between user c and user q,
Indicate user q to the explicit degree of belief of user d;
Step 4: the prediction degree of belief between user generated according to step 3 calculates target user to project using formula (4)
Prediction scoring;
Wherein,It is in explicit trusting relationship, user c scores to the prediction of project i,It is user c and user d
Between prediction degree of belief,It is score value of the user d to project i,Indicate the number of users in users to trust rating matrix
Amount;
Step 5: according to user-project rating matrix, calculating the implicit trust degree between user using formula (5), obtain user
Implicit trust matrix,
Wherein,It is the implicit trust degree between user c and user d,It is score value of the user c to project i,
Indicate that user c comments excessive all items,It indicates that user d had project i and is selected as 1, do not selected to be 0,
Indicate that user d comments the number of excessive project;
Step 6: the user concealed trust matrix obtained according to step 5 calculates target user to the pre- of project using formula (6)
Assessment point;
Wherein,It is in implicit trust relationship, user c scores to the prediction of project i,It is user c and user d
Between implicit trust degree,It is score value of the user d to project i,Indicate the number of users in user-project rating matrix
Amount;
Step 7: for root according to step 4 and step 6 as a result, being merged using formula (7), prediction target user comments project
Score value, and descending arrangement is carried out, prediction is scored highest K project recommendation to target user, generates recommendation list;
Wherein,It is prediction score value of the target user c to project i,It is fusion factor, its value range is (0,1),It is in explicit trusting relationship, user c scores to the prediction of project i,It is in implicit trust relationship, user c is to project i's
Prediction scoring.
Claims (1)
1. a kind of based on the explicit personalized recommendation method trusted with implicit trust, it is characterised in that: assuming that existing subscriber-project
Rating matrix and users to trust rating matrix, include the following steps:
Step 1: according to users to trust rating matrix, the trust similarity between user being calculated using formula (1), and by drop
Sequence arrangement is chosen and trusts neighbours' collection recently with the highest top n user composition of user c trust similitude;
Wherein,Indicate the trust similarity between user c and user d,It is user c and user d
The user's number trusted jointly,It is the sum of user in users to trust rating matrix;
Step 2: trusting neighbours' collection recently according to obtained in step 1, calculated using formula (2) and trust neighbours concentration user recently
Between explicit degree of belief;
Wherein,Indicate user c to the explicit degree of belief of user d,Indicate the in-degree of user d,Table
Show maximum in-degree in all users of user c trust;
Step 3: the explicit degree of belief between the user according to obtained in step 2 calculates the prediction between user using formula (3)
Degree of belief;
Wherein,It is the prediction degree of belief between user c and user d,It is and user's c similitude highest
Top n trust recently neighbours collection,It is the trust similarity between user c and user q,Table
Show user q to the explicit degree of belief of user d;
Step 4: the prediction degree of belief between user generated according to step 3 calculates target user to project using formula (4)
Prediction scoring;
Wherein,It is in explicit trusting relationship, user c scores to the prediction of project i,It is user c and user d
Between prediction degree of belief,It is score value of the user d to project i,Indicate the number of users in users to trust rating matrix
Amount;
Step 5: according to user-project rating matrix, calculating the implicit trust degree between user using formula (5), obtain user
Implicit trust matrix,
Wherein,It is the implicit trust degree between user c and user d,It is score value of the user c to project i,
Indicate that user c comments excessive all items,It indicates that user d had project i and is selected as 1, do not selected to be 0,
Indicate that user d comments the number of excessive project;
Step 6: the user concealed trust matrix obtained according to step 5 calculates target user to the pre- of project using formula (6)
Assessment point;
Wherein,It is in implicit trust relationship, user c scores to the prediction of project i,It is user c and user d
Between implicit trust degree,It is score value of the user d to project i,Indicate the number of users in user-project rating matrix
Amount;
Step 7: for root according to step 4 and step 6 as a result, being merged using formula (7), prediction target user comments project
Score value, and descending arrangement is carried out, prediction is scored highest K project recommendation to target user, generates recommendation list;
Wherein,It is prediction score value of the target user c to project i,It is fusion factor, its value range is (0,1),
It is in explicit trusting relationship, user c scores to the prediction of project i,It is in implicit trust relationship, user c is pre- to project i's
Assessment point.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811152650.2A CN109101667B (en) | 2018-09-29 | 2018-09-29 | Personalized recommendation method based on explicit trust and implicit trust |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811152650.2A CN109101667B (en) | 2018-09-29 | 2018-09-29 | Personalized recommendation method based on explicit trust and implicit trust |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109101667A true CN109101667A (en) | 2018-12-28 |
CN109101667B CN109101667B (en) | 2021-07-09 |
Family
ID=64868149
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811152650.2A Active CN109101667B (en) | 2018-09-29 | 2018-09-29 | Personalized recommendation method based on explicit trust and implicit trust |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109101667B (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109933721A (en) * | 2019-02-01 | 2019-06-25 | 中森云链(成都)科技有限责任公司 | A kind of interpretable recommended method merging user concealed article preference and implicit trust |
CN109948067A (en) * | 2019-02-22 | 2019-06-28 | 哈尔滨工业大学(深圳) | The information-pushing method and system of user's enigmatic language justice LR model are trusted in a kind of fusion |
CN110334286A (en) * | 2019-07-10 | 2019-10-15 | 南京工业大学 | A kind of personalized recommendation method based on trusting relationship |
CN111460318A (en) * | 2020-03-31 | 2020-07-28 | 中南大学 | Collaborative filtering recommendation method based on explicit and implicit trusts |
Citations (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20130097056A1 (en) * | 2011-10-13 | 2013-04-18 | Xerox Corporation | Methods and systems for recommending services based on an electronic social media trust model |
CN103995823A (en) * | 2014-03-25 | 2014-08-20 | 南京邮电大学 | Information recommending method based on social network |
CN104915391A (en) * | 2015-05-25 | 2015-09-16 | 南京邮电大学 | Article recommendation method based on trust relationship |
CN105279699A (en) * | 2015-10-09 | 2016-01-27 | 北京航空航天大学 | Recommendation method combining multi-class untrust relation based on collaborative filtering |
CN105354260A (en) * | 2015-10-22 | 2016-02-24 | 中南大学 | Mobile application recommendation method with social network and project feature fused |
CN105809510A (en) * | 2016-03-04 | 2016-07-27 | 王瑞琴 | Multi-faceted social trust based collaborative recommendation method |
CN106354783A (en) * | 2016-08-23 | 2017-01-25 | 武汉大学 | Social recommendation method based on trust relationship implicit similarity |
CN106570090A (en) * | 2016-10-20 | 2017-04-19 | 杭州电子科技大学 | Method for collaborative filtering recommendation based on interest changes and trust relations |
CN106682114A (en) * | 2016-12-07 | 2017-05-17 | 广东工业大学 | Personalized recommending method fused with user trust relationships and comment information |
CN107025606A (en) * | 2017-03-29 | 2017-08-08 | 西安电子科技大学 | The item recommendation method of score data and trusting relationship is combined in a kind of social networks |
CN107301583A (en) * | 2017-05-26 | 2017-10-27 | 重庆邮电大学 | It is a kind of that method is recommended based on user preference and the cold start-up trusted |
CN107330461A (en) * | 2017-06-27 | 2017-11-07 | 安徽师范大学 | Collaborative filtering recommending method based on emotion with trust |
-
2018
- 2018-09-29 CN CN201811152650.2A patent/CN109101667B/en active Active
Patent Citations (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20130097056A1 (en) * | 2011-10-13 | 2013-04-18 | Xerox Corporation | Methods and systems for recommending services based on an electronic social media trust model |
CN103995823A (en) * | 2014-03-25 | 2014-08-20 | 南京邮电大学 | Information recommending method based on social network |
CN104915391A (en) * | 2015-05-25 | 2015-09-16 | 南京邮电大学 | Article recommendation method based on trust relationship |
CN105279699A (en) * | 2015-10-09 | 2016-01-27 | 北京航空航天大学 | Recommendation method combining multi-class untrust relation based on collaborative filtering |
CN105354260A (en) * | 2015-10-22 | 2016-02-24 | 中南大学 | Mobile application recommendation method with social network and project feature fused |
CN105809510A (en) * | 2016-03-04 | 2016-07-27 | 王瑞琴 | Multi-faceted social trust based collaborative recommendation method |
CN106354783A (en) * | 2016-08-23 | 2017-01-25 | 武汉大学 | Social recommendation method based on trust relationship implicit similarity |
CN106570090A (en) * | 2016-10-20 | 2017-04-19 | 杭州电子科技大学 | Method for collaborative filtering recommendation based on interest changes and trust relations |
CN106682114A (en) * | 2016-12-07 | 2017-05-17 | 广东工业大学 | Personalized recommending method fused with user trust relationships and comment information |
CN107025606A (en) * | 2017-03-29 | 2017-08-08 | 西安电子科技大学 | The item recommendation method of score data and trusting relationship is combined in a kind of social networks |
CN107301583A (en) * | 2017-05-26 | 2017-10-27 | 重庆邮电大学 | It is a kind of that method is recommended based on user preference and the cold start-up trusted |
CN107330461A (en) * | 2017-06-27 | 2017-11-07 | 安徽师范大学 | Collaborative filtering recommending method based on emotion with trust |
Non-Patent Citations (5)
Title |
---|
GUIBING GUO: ""From ratings to trust:an empirical study of implicit trust in recommender systems"", 《SAC "14: PROCEEDINGS OF THE 29TH ANNUAL ACM SYMPOSIUM ON APPLIED COMPUTING》 * |
GUIBING GUO: ""Trustsvd Collaborative filtering with both the explicit and implicit influence of user trust and of item ratings"", 《PROCEEDINGS OF THE TWENTY-NINTH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE》 * |
SWATI GUPTA: ""Trust Aware Recommender Systems A Survey on Implicit Trust Generation Techniques"", 《INTERNATIONAL JOURNAL OF COMPUTER SCIENCE AND INFORMATION TECHNOLOGIES》 * |
贾冬艳: ""基于双重邻居选取策略的协同过滤推荐算法"", 《计算机研究与发展》 * |
陆坤: ""一种融合隐式信任的协同过滤推荐算法"", 《小型微型计算机系统》 * |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109933721A (en) * | 2019-02-01 | 2019-06-25 | 中森云链(成都)科技有限责任公司 | A kind of interpretable recommended method merging user concealed article preference and implicit trust |
CN109948067A (en) * | 2019-02-22 | 2019-06-28 | 哈尔滨工业大学(深圳) | The information-pushing method and system of user's enigmatic language justice LR model are trusted in a kind of fusion |
CN110334286A (en) * | 2019-07-10 | 2019-10-15 | 南京工业大学 | A kind of personalized recommendation method based on trusting relationship |
CN111460318A (en) * | 2020-03-31 | 2020-07-28 | 中南大学 | Collaborative filtering recommendation method based on explicit and implicit trusts |
CN111460318B (en) * | 2020-03-31 | 2022-09-30 | 中南大学 | Collaborative filtering recommendation method based on explicit and implicit trusts |
Also Published As
Publication number | Publication date |
---|---|
CN109101667B (en) | 2021-07-09 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109101667A (en) | A kind of personalized recommendation method based on explicit trust and implicit trust | |
Zheng et al. | Deep‐learning contact‐map guided protein structure prediction in CASP13 | |
Shi et al. | Heterogeneous information network embedding for recommendation | |
Zhang et al. | Template‐based and free modeling of I‐TASSER and QUARK pipelines using predicted contact maps in CASP12 | |
Zhang et al. | Negative-aware attention framework for image-text matching | |
Ginalski et al. | ORFeus: detection of distant homology using sequence profiles and predicted secondary structure | |
US9792640B2 (en) | Generating and providing content recommendations to a group of users | |
CN109241203B (en) | Clustering method for user preference and distance weighting by fusing time factors | |
CN109933721B (en) | Interpretable recommendation method integrating user implicit article preference and implicit trust | |
WO2018121700A1 (en) | Method and device for recommending application information based on installed application, terminal device, and storage medium | |
Wang et al. | A multi-task learning approach for improving product title compression with user search log data | |
CN101976148B (en) | Hand input system and method | |
US9268767B2 (en) | Semantic-based search system and search method thereof | |
US20160283952A1 (en) | Ranking information providers | |
CN103729359A (en) | Method and system for recommending search terms | |
JP2015511363A (en) | Determination of search result ranking based on confidence level values associated with sellers | |
CN112966009A (en) | Venation knowledge panel | |
WO2016202214A2 (en) | Method and device for displaying keyword | |
CN107025606B (en) | Project recommendation method combining evaluation data and trust relationship in social network | |
CN109376301A (en) | A kind of personalized recommendation method merging social information | |
US20170323019A1 (en) | Ranking information providers | |
CN113961823B (en) | News recommendation method, system, storage medium and equipment | |
US11341141B2 (en) | Search system using multiple search streams | |
CN109190045A (en) | A kind of personalized recommendation method based on bigraph (bipartite graph) arest neighbors | |
AU2013331585B2 (en) | System, method and interface for providing a search result using segment constraints |
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 |