CN103309976A - Method for improving social recommendation efficiency based on user personality - Google Patents
Method for improving social recommendation efficiency based on user personality Download PDFInfo
- Publication number
- CN103309976A CN103309976A CN2013102314494A CN201310231449A CN103309976A CN 103309976 A CN103309976 A CN 103309976A CN 2013102314494 A CN2013102314494 A CN 2013102314494A CN 201310231449 A CN201310231449 A CN 201310231449A CN 103309976 A CN103309976 A CN 103309976A
- Authority
- CN
- China
- Prior art keywords
- commodity
- recommendation
- score
- targeted customer
- user
- 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
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a method for improving the social recommendation efficiency based on user personality. The method comprises the steps as follows: a, determining k commodities to be recommended to a target user; b, for each commodity m to be recommended, calculating the comprehensive recommendation scores of the good friends of the target user based on personalities and commodity preferences respectively, and selecting the friend with the highest score as a corresponding recommender of the commodity m; and c, notifying each recommender to recommend the corresponding commodity to the target user. According to the method, the proper recommender is selected according to the user personality features and the commodity preferences, so that the recommendation efficiency is improved; and the method has the advantages as follows: the loss in a recommendation transfer process can be further reduced, the circulation of the commodities in a social network can be further promoted, and the social recommendation efficiency can be further improved.
Description
Technical field
The present invention relates to social activity recommendation method is improved to improve the commending system research field of its efficient, specifically by in targeted customer's good friend, seek on the psychological characteristics (referring to personality) and be willing to that more the people of others' Recommendations of purpose is as the recommendation people, improve the possibility that commodity arrive the targeted customer, thereby improve the efficient of social recommendation.
Background technology
In the research that relevant socialization is recommended, the researcher often concentrates on the social relationships that how to utilize the user and (appears in the system recommendation tabulation as commodity m for recommendation results provides explanation, meeting supplemental instruction m is that some good friend likes), to increase user's degree of belief.Present stage, difference between this dual mode of friend recommendation is directly recommended and utilized in some researchist system that begins one's study.
Studies show that in a large number compare with the commodity of system recommendation, the user more has a preference for the recommendation from the good friend.That is to say that even same commodity, source channel difference is respectively system recommendation and good friend recommendation, the user is significantly higher than the former to the latter's good opinion.Therefore, some systems have additionally increased the way of recommendation of this good friend's recommendation, to obtain higher commodity receptance.Yet, the mode efficient of present this good friend's recommendation is also very low, main cause has been to ignore the recommendation wish of good friend self (following general designation recommendation people), the recommendation people who namely chooses is very willing to dependent merchandise is passed to social networks, the mode that causes the efficient under the recommendation mode directly to recommend far below system.
On the other hand, can influence the theory of behavior according to the relevant personality in psychological study field, that nature factor is included at present a lot of commending systems, effectively improved the recommendation performance.Therefore perhaps can adopt similar mode to alleviate the social activity of being unwilling to recommend to cause because of the recommendation people and recommend this problem of inefficiency.
Summary of the invention
The objective of the invention is at having ignored in the prior art that the recommendation people recommends the technological deficiency of wish and a kind of method that improves social recommendation efficient based on user's personality of providing, this method more is ready to participate in the recommendation people that commodity transmit in character by seeking in all good friends of targeted customer, notify these recommendation people to give the targeted customer related products recommendation, improve the efficient of social recommendation with transmission loss still less.
A kind of method based on the social recommendation of user's personality raising efficient, this method may further comprise the steps:
A, definite k commodity will recommending the targeted customer specifically comprise:
ⅰ) find the targeted customer not have the commodity set A of consumer behavior;
ⅱ) for each commodity in the commodity set A, use the collaborative filtering based on the user, dope the targeted customer to the hobby score of these commodity;
ⅲ) all commodity in the commodity set A are sorted from big to small by prediction hobby score, choose the highest k of a score commodity, be the targeted customer's of giving to be recommended commodity;
B, for each commodity m to be recommended, calculate each good friend of targeted customer respectively based on the comprehensive recommendation score of personality and commodity hobby, choose score the highest recommend the people accordingly as commodity m, specifically comprise:
ⅰ) from the user's personality database based on " five-factor model personality " model (big-five model), search each good friend of targeted customer at the score s_p of this personality dimension of " pleasant property " (being agreeableness in the relevant english literature);
ⅱ) according to user-commodity score data storehouse, obtain each good friend to the hobby of commodity m scoring s_r(reality or prediction);
ⅲ) according to step ⅰ) and ⅱ), calculate each good friend's comprehensive recommendation score:
Comprehensive recommendation score
ⅳ) each good friend is sorted from big to small according to comprehensive recommendation score, what come first recommends the people accordingly as commodity m;
C, notice respectively recommend the people that corresponding commodity are recommended to the targeted customer, specifically comprise:
ⅰ) send message to the recommendation people, message content comprises dependent merchandise and the corresponding targeted customer that will recommend;
ⅱ) determine whether recommend by the recommendation people.
The ⅱ of described step b) specifically comprise:
1., if this good friend is arranged to the scoring record of commodity m in user-commodity score data storehouse, then should scoring record and be the good friend to the hobby of the commodity m s_r that marks, 2. calculate otherwise carry out step;
2., use collaborative filtering based on the user calculate this good friend to the prediction scoring of commodity m as s_r.
Compare with background technology, the present invention has following advantage:
The present invention is carrying out social activity when recommendation, mainly from recommendation people's angle, taken all factors into consideration recommendation people's character trait and to the hobby situation of commodity.
Why consider character trait, associated user's experiment of doing before mainly being based on." pleasant property " personality dimension (based on " five-factor model personality " model) user that score is more high is found in experiment, more other people Recommendations of hope purpose.According to psychology field available research achievements, " pleasant property " dimension score is more high, and the expression user more likes cooperating with people, and is for the people is honest, ready to help others; Otherwise then representative is more selfish, likes suspecting other people.Therefore experiment is found to match with available research achievements, can consider to filter out suitable recommendation people with character trait, and minimizing is unwilling to recommend the transmission loss that causes because of the recommendation people.
The present invention has considered that recommendation people self treats the hobby situation of recommendation commodity.Usually the user more is ready to recommend oneself to feel the measured commodity of matter.Therefore the hobby of considering recommendation people self can further improve the recommendation wish, thereby improves the efficient of social recommendation.
In addition, the commodity of recommending the targeted customer are to use based on user's collaborative filtering to choose out, meet targeted customer's hobby, recommendation be the interested commodity of targeted customer's most probable, also be beneficial to the targeted customer and accept, recommend more effective.
For the targeted customer who does not have the good friend, will not take social recommendation method, the mode that can adopt direct recommendation or authoritative expert to recommend.
Description of drawings
Fig. 1 is schematic flow sheet of the present invention.
Fig. 2 is good friend's character trait and the commodity hobby synoptic diagram in the embodiment of the invention.
Embodiment
Consult Fig. 1, the present invention is applicable in the social activity recommendation of all commodity.At first determine to recommend targeted customer's k commodity with related algorithm, then for each commodity that will recommend, according to comprehensive recommendation score, in targeted customer's good friend, choose only recommendation people, notice respectively recommends the people to give the targeted customer with corresponding commercial product recommending at last, and its concrete steps are as follows:
(1): find the targeted customer not have the commodity set A of consumer behavior, for each commodity among the A, use the collaborative filtering based on the user, dope the targeted customer to the hobby score of these commodity, select the highest k of score, as the commodity of giving the targeted customer to be recommended;
(2): choose a commodity m in the k from step (1) the commodity to be recommended at every turn;
(3): all good friends that obtain the targeted customer gather F;
(4): judge good friend's number | whether F| greater than 0, if | F|〉0, then change step (5) over to, otherwise change end over to;
(5): i. is according to user-commodity score data storehouse, obtain each good friend to the hobby of commodity m scoring s_r(reality or prediction), ii. from the user's personality database based on " five-factor model personality " model, search each good friend at the score s_p of personality dimension " pleasant property ";
(6): by following formula, calculate each good friend to the comprehensive recommendation score of commodity m:
Then all good friends are sorted from high to low by score;
(7): choose the highest good friend of comprehensive recommendation score as the recommendation people of commodity m correspondence;
(8): notify this recommendation people that commodity m is recommended the targeted customer;
(9): judge whether to have finished the recommendation work to all commodity (k that has selected), "Yes" then changes end over to, and "No" then changes step (2) over to.
Embodiment
The embodiment that recommends by following film understands the present invention better.
Suppose will recommend film to certain targeted customer with the mode (namely passing through the good friend) of social recommendation now, its concrete steps are as follows:
(1) by based on user's collaborative filtering, the film of not seen the targeted customer is concentrated and selected the k(that will recommend is 2 herein) portion's film is respectively " safe Embarrassing ", " the strange dog of science ";
(2) select a wherein film (as " safe Embarrassing ") conduct film to be recommended;
(3) obtain targeted customer all good friends in this system and gather F={f
1,f
2,f
3;
(4) because good friend's number | F|〉0, change step (5) over to;
(5) obtain each good friend the score s_p of personality dimension " pleasant property " and to the hobby scoring s_r(reality of this film or prediction), see Fig. 2;
(6) pass through formula
Calculate each good friend to the comprehensive recommendation score of this film:
S(f
1, " safe Embarrassing ")=1.597;
S(f
2, " safe Embarrassing ")=1.766;
S(f
3, " safe Embarrassing ")=1.771;
Each good friend sorts from high to low by score and is f
3, f
2, f
1
(7) comprehensively recommend the highest good friend f of score
3Be the corresponding recommendation people of film " safe Embarrassing ";
(8) send out message informing f
3" safe Embarrassing " recommends this targeted customer with film;
(9) by judging, also do not finish the recommendation work to this k portion film, need change step (2) over to, recommend film " the strange dog of science " with similar method to the targeted customer.
Claims (2)
1. one kind is improved the method for social recommendation efficient based on user's personality, it is characterized in that this method may further comprise the steps:
A, definite k commodity will recommending the targeted customer specifically comprise:
ⅰ) find the targeted customer not have the commodity set A of consumer behavior;
ⅱ) for each commodity in the commodity set A, use the collaborative filtering based on the user, dope the targeted customer to the hobby score of these commodity;
ⅲ) all commodity in the commodity set A are sorted from big to small by prediction hobby score, choose the highest k of a score commodity, be the targeted customer's of giving to be recommended commodity;
B, for each commodity m to be recommended, calculate each good friend of targeted customer respectively based on the comprehensive recommendation score of personality and commodity hobby, choose score the highest recommend the people accordingly as commodity m, specifically comprise:
ⅰ) from the user's personality database based on " five-factor model personality " model, search each good friend of targeted customer at the score s_p of " pleasant property " this personality dimension;
ⅱ) according to user-commodity score data storehouse, obtain each good friend to the hobby scoring s_r of commodity m;
ⅲ) according to step ⅰ) and ⅱ), calculate each good friend's comprehensive recommendation score:
Comprehensive recommendation score
ⅳ) each good friend is sorted from big to small according to comprehensive recommendation score, what come first recommends the people accordingly as commodity m;
C, notice respectively recommend the people that corresponding commodity are recommended to the targeted customer, specifically comprise:
ⅰ) send message to the recommendation people, message content comprises dependent merchandise and the corresponding targeted customer that will recommend;
ⅱ) determine whether recommend by the recommendation people.
2. according to claim 1 based on the social method of recommending efficient of user's personality raising, it is characterized in that the ⅱ of described step b) specifically comprise:
1., if this good friend is arranged to the scoring record of commodity m in user-commodity score data storehouse, then should scoring record and be the good friend to the hobby of the commodity m s_r that marks, 2. calculate otherwise carry out step;
2., use collaborative filtering based on the user calculate this good friend to the prediction scoring of commodity m as s_r.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310231449.4A CN103309976B (en) | 2013-06-13 | 2013-06-13 | Method for improving social recommendation efficiency based on user personality |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310231449.4A CN103309976B (en) | 2013-06-13 | 2013-06-13 | Method for improving social recommendation efficiency based on user personality |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103309976A true CN103309976A (en) | 2013-09-18 |
| CN103309976B CN103309976B (en) | 2017-02-08 |
Family
ID=49135194
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310231449.4A Expired - Fee Related CN103309976B (en) | 2013-06-13 | 2013-06-13 | Method for improving social recommendation efficiency based on user personality |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103309976B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104008141A (en) * | 2014-05-08 | 2014-08-27 | 南京邮电大学 | Product grading method by user based on on-line social network |
| CN105282702A (en) * | 2015-09-07 | 2016-01-27 | 广东欧珀移动通信有限公司 | Indoor positioning method and user terminal |
| CN108109058A (en) * | 2018-01-11 | 2018-06-01 | 合肥工业大学 | A kind of single classification collaborative filtering method for merging personal traits and article tag |
| CN108647216A (en) * | 2017-03-16 | 2018-10-12 | 上海交通大学 | Software crowdsourcing task recommendation system and method based on developer's social networks |
| CN109598578A (en) * | 2018-11-09 | 2019-04-09 | 深圳壹账通智能科技有限公司 | The method for pushing and device of business object data, storage medium, computer equipment |
| EP3675001A1 (en) | 2018-12-27 | 2020-07-01 | Telefonica Innovacion Alpha S.L | A computer implemented method, a system and computer program for determining optimal behavior path for a user |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102880691A (en) * | 2012-09-19 | 2013-01-16 | 北京航空航天大学深圳研究院 | User closeness-based mixed recommending system and method |
| CN102968506A (en) * | 2012-12-14 | 2013-03-13 | 北京理工大学 | Personalized collaborative filtering recommendation method based on extension characteristic vectors |
-
2013
- 2013-06-13 CN CN201310231449.4A patent/CN103309976B/en not_active Expired - Fee Related
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102880691A (en) * | 2012-09-19 | 2013-01-16 | 北京航空航天大学深圳研究院 | User closeness-based mixed recommending system and method |
| CN102968506A (en) * | 2012-12-14 | 2013-03-13 | 北京理工大学 | Personalized collaborative filtering recommendation method based on extension characteristic vectors |
Non-Patent Citations (3)
| Title |
|---|
| MOHAMMAD S AMIN，ETC.: "Social Referral: Leveraging Network Connections to Deliver Recommendations", 《RECSYS’12》, 13 September 2012 (2012-09-13), pages 273 - 276 * |
| WEN WU,ETC.: "Using Personality to Adjust Diversity in Recommender Systems", 《PROC.HT2013,ACM PRESS》, 3 May 2013 (2013-05-03), pages 225 - 229 * |
| 李晓娥: "SNS社交网站信息分享行为的影响因素", 《媒体时代》, 15 April 2011 (2011-04-15), pages 18 - 21 * |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104008141A (en) * | 2014-05-08 | 2014-08-27 | 南京邮电大学 | Product grading method by user based on on-line social network |
| CN105282702A (en) * | 2015-09-07 | 2016-01-27 | 广东欧珀移动通信有限公司 | Indoor positioning method and user terminal |
| CN108647216A (en) * | 2017-03-16 | 2018-10-12 | 上海交通大学 | Software crowdsourcing task recommendation system and method based on developer's social networks |
| CN108109058A (en) * | 2018-01-11 | 2018-06-01 | 合肥工业大学 | A kind of single classification collaborative filtering method for merging personal traits and article tag |
| CN108109058B (en) * | 2018-01-11 | 2021-06-29 | 合肥工业大学 | Single-classification collaborative filtering method integrating personality traits and article labels |
| CN109598578A (en) * | 2018-11-09 | 2019-04-09 | 深圳壹账通智能科技有限公司 | The method for pushing and device of business object data, storage medium, computer equipment |
| EP3675001A1 (en) | 2018-12-27 | 2020-07-01 | Telefonica Innovacion Alpha S.L | A computer implemented method, a system and computer program for determining optimal behavior path for a user |
| WO2020135936A1 (en) | 2018-12-27 | 2020-07-02 | Telefonica Innovación Alpha S.L | A computer implemented method, a system and computer program for determining personalized parameters for a user |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103309976B (en) | 2017-02-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103309976A (en) | Method for improving social recommendation efficiency based on user personality | |
| CN102999507B (en) | The recommendation process method and apparatus of network microblog famous person's information | |
| CN101685458A (en) | Recommendation method and system based on collaborative filtering | |
| Joshi et al. | Performance evaluation of agro-tourism clusters using AHP–TOPSIS | |
| Gellert | Neoliberalism and altered state developmentalism in the twenty-first century extractive regime of Indonesia | |
| US11663661B2 (en) | Apparatus and method for training a similarity model used to predict similarity between items | |
| JP6015959B2 (en) | Information processing apparatus, information processing method, and program | |
| DE212010000172U1 (en) | Social search engine | |
| SG171594A1 (en) | Website management method and on-line system | |
| CN106407349A (en) | Product recommendation method and device | |
| CN108876069A (en) | A kind of endowment service recommendation method | |
| JP2019507425A (en) | Service processing method, data processing method and apparatus | |
| CN103970801B (en) | Microblogging advertisement blog article recognition methods and device | |
| US20140032332A1 (en) | Promoting products on a social networking system based on information from a merchant site | |
| Chaurasiya et al. | Improving performance of product recommendations using user reviews | |
| CN104915391A (en) | Article recommendation method based on trust relationship | |
| KR101381921B1 (en) | Apparatus and method for classifying propensity of consumer | |
| Bhojne et al. | Collaborative approach based restaurant recommender system using Naive Bayes | |
| CN105279180A (en) | Two-way selection based recommendation framework | |
| CN105208033A (en) | Group auxiliary recommendation method and system based on intelligent terminal scenes | |
| Kobyashi et al. | Non-task-oriented dialogue system considering user’s preference and human relations | |
| CN113362034A (en) | Position recommendation method | |
| Li et al. | Effect analysis of tiktok gourmet bloggers’ videos for social media marketing | |
| CN108665306B (en) | Core competition product identification method and system, and storage medium | |
| JP2012221360A (en) | System capable of providing information deliverer (influencer) influential for purchasing behavior with advertisement information |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20170208 Termination date: 20180613 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |