CN108241619A - A kind of recommendation method based on the more interest of user - Google Patents

A kind of recommendation method based on the more interest of user Download PDF

Info

Publication number
CN108241619A
CN108241619A CN201611203999.5A CN201611203999A CN108241619A CN 108241619 A CN108241619 A CN 108241619A CN 201611203999 A CN201611203999 A CN 201611203999A CN 108241619 A CN108241619 A CN 108241619A
Authority
CN
China
Prior art keywords
user
interest
article
score
value
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN201611203999.5A
Other languages
Chinese (zh)
Inventor
田小伟
吴江
马力
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.)
Northwest University
Original Assignee
Northwest University
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 Northwest University filed Critical Northwest University
Priority to CN201611203999.5A priority Critical patent/CN108241619A/en
Publication of CN108241619A publication Critical patent/CN108241619A/en
Pending legal-status Critical Current

Links

Classifications

    • 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]
    • G06Q30/0631—Item recommendations

Landscapes

  • Engineering & Computer Science (AREA)
  • Databases & Information Systems (AREA)
  • Theoretical Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Accounting & Taxation (AREA)
  • Finance (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Development Economics (AREA)
  • Economics (AREA)
  • Marketing (AREA)
  • Strategic Management (AREA)
  • General Business, Economics & Management (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

The invention discloses a kind of recommendation methods based on the more interest of user, it is characterised in that：During target user's trusted neighbor user is determined, neighbours' screening technique of the more attribute fusions of user is proposed；According to the data to have scored in user items, the more interest portraits of structure user；It proposes to draw a portrait based on more interest and the score in predicting method of similitude weighting is merged with more attributes；It proposes a kind of more interest quaternary group models of user, is compared with conventional recommendation, the characteristics of recommendation results of the present invention more agree with user's more interest, recommendation results have diversity；User's more attributes fusion similarity neighbor choices in some degree, solve the problems, such as cold start-up；Compare with traditional collaborative filtering, accuracy rate improves.

Description

A kind of recommendation method based on the more interest of user
Technical field
The present invention relates to a kind of recommendation methods based on the more interest of user.
Background technology
Recommendation refers to recommend possible interested product for user according to the data of user and behavior.Personalized recommendation is mutual One important field of research of the Internet services epoch.At present, recommended technology is widely used to the crowds such as e-commerce, social networks It is multi-field.The development of information technology, the rapid growth of data, make conventional recommendation algorithm face huge challenge.The diversification of interest The fact that be a generally existing.And the diversification of people's interest is determined in proposed algorithm field, the diversification of recommendation results It is very important.Herein as point of penetration, the more interest proposed algorithms of user are studied.Under normal conditions, recommendation results are often It is that the high article that scores is recommended into user or the article higher with user preferences article similarity is recommended into user. However, in many field scenes, the demand of user is to recommend the article preferred in category as much as possible, that is, meet More interest of user.
It is that traditional proposed algorithm is recommended the result is that often single interest.Traditional proposed algorithm is to according to searching target The higher article of the neighbor user interest-degree of user, that is to say, that acquiescence has arranged neighbor user and target user have it is identical Hobby interests.In this case, if the neighbor user interest of target user is single, target user's interest is extensive, that basis This neighbor user predicts the hobby of the target user, and error can be bigger.
Invention content
It is an object of the invention to overcome above-mentioned deficiency in the prior art, and providing one kind can realize that user is more Interest is recommended and the recommendation method based on the more interest of user of more interest recommendation functions of the more attributes of user and classification of the items.
Technical solution is used by the present invention solves the above problems：Based on the recommendation method of the more interest of user, feature It is：During target user's trusted neighbor user is determined, neighbours' screening technique of the more attribute fusions of user is proposed；According to The data to have scored in user-project, the more interest portraits of structure user；It proposes to merge with more attributes based on more interest portrait similar Property weighting score in predicting method；It proposes a kind of more interest quaternary group models of user, is as follows：
Step 1. user is more, and interest model is expressed as：(R, I, C, P) four-tuple, R are consumer articles scoring set, and I is represented User-article hobby set, C represent article-category set, and P represents user-category set；
Step 2.R gathers for consumer articles scoring, R=(R1,R2,…,Rn)T, n indicates n user, if 0<i<=n, Ri Represent user's score information of i-th of user, each user's score information table is made of many scorings to article, Ri=(Ri1, Ri2,…,Rim), m represents m article.R in matrix RijNumerical value represent user uiTo article vjScoring, note score value be Score, ordinary circumstance agreement 1<=score<=5,1 representative does not like, and 5 representatives enjoy a lot, and -1 representative is not scored；
Step 3.I represents user-article hobby set, Iil(1<=i<=n, 1<=l<=j) numerical value represent hobby journey Degree, numerical value is bigger, and expression is more liked.Its value represents not liking for 0 or 1,0, and 1 represents to like；
Step 4.C represents article-category set, represents the relationship of article and goods categories, and the value of element indicates whether It is affiliated such, belong to value for 1, it is 0 to be not belonging to value；
Step 5.P represents user-category set, reflects the relationship between user and goods categories M.Namely user draws Picture；
Step 6. data prediction obtains user's essential attribute information and the user items matrix that scored；Remember that user collects to use Family set U={ U1,U2,U3,U4,U5,Um, article set V={ V1,V2,V3…VnConsumer articles grade form matrix Rm*nTable Show, r belongs to Rm*n, rmnRepresent user Um to article VnScoring；
The more attribute fusion similarity calculation Msim (U of step 7. useri,Vj)；The personal essential information of user, generally comprises： Gender, age, educational background, occupation, hobby etc.；User information is expressed as (X1, X2, X3, X4, X5, Xk), k represents user property Number；According to different attribute feature, suitable similarity calculating method is selected respectively；So, Msim (Ui,Vj)=a1sim (U1X1,V2X1)+a2sim(U1X2,V2X2)+…+aksim(U1Xk,V2Xk)；
Step 8. builds user's trusted neighborhood；Neighbours number max-thresholds P is set, is merged according to the more attributes of user Similarity calculation is as a result, obtain the neighbor user that target user is not more than P；
Step 9. builds project-item class table, and according to user items rating matrix, by mapping relations, it is more to obtain user Interest is drawn a portrait；
The more interest portraits of trusted neighborhood and step 9 user that step 10. is obtained with step 8 for constraint, use by filling Family-project rating matrix obtains non-sparse rating matrix；
Step 11. project high to the score value of target user does Top-N sequences, as recommendation results.
Compared with prior art, the present invention haing the following advantages and effect：1) it is compared with conventional recommendation, recommendation results are more The characteristics of agreeing with user's more interest, recommendation results have diversity；2) the more attribute fusion similarity neighbor choices of user, some journeys On degree, solves the problems, such as cold start-up；3) compare with Slope One and traditional collaborative filtering, accuracy rate improves.
Description of the drawings
Fig. 1 is the flow principle schematic of the embodiment of the present invention.
Fig. 2 is the flow diagram of the specific implementation step of the embodiment of the present invention.
Specific embodiment
Below in conjunction with the accompanying drawings and pass through embodiment the present invention is described in further detail, following embodiment is to this hair Bright explanation and the invention is not limited in following embodiments.Step is as follows：
Referring to Fig. 1, based on the recommendation method of the more interest of user, during target user's trusted neighbor user is determined, It is proposed neighbours' screening technique of the more attribute fusions of user；According to the data to have scored in user-project, the more interest of structure user are drawn Picture；It proposes to draw a portrait based on more interest and the score in predicting method of similitude weighting is merged with more attributes；It is proposed a kind of more interest of user Quaternary group model,
As shown in Fig. 2, it is as follows：
Step 1. user is more, and interest model is expressed as：(R, I, C, P) four-tuple, R are consumer articles scoring set, and I is represented User-article hobby set, C represent article-category set, and P represents user-category set；
Step 2.R gathers for consumer articles scoring, R=(R1,R2,…,Rn)T, n indicates n user, if 0<i<=n, Ri Represent user's score information of i-th of user, each user's score information table is made of many scorings to article, Ri=(Ri1, Ri2,…,Rim), m represents m article.R in matrix RijNumerical value represent user uiTo article vjScoring, note score value be Score, ordinary circumstance agreement 1<=score<=5,1 representative does not like, and 5 representatives enjoy a lot, and -1 representative is not scored；
Step 3.I represents user-article hobby set, Iil(1<=i<=n, 1<=l<=j) numerical value represent hobby journey Degree, numerical value is bigger, and expression is more liked.Its value represents not liking for 0 or 1,0, and 1 represents to like；
Step 4.C represents article-category set, represents the relationship of article and goods categories, and the value of element indicates whether It is affiliated such, belong to value for 1, it is 0 to be not belonging to value；
Step 5.P represents user-category set, reflects the relationship between user and goods categories M.Namely user draws Picture；
Step 6. data prediction obtains user's essential attribute information and the user items matrix that scored；Remember that user collects to use Family set U={ U1,U2,U3,U4,U5,Um, article set V={ V1,V2,V3…VnConsumer articles grade form matrix Rm*nTable Show, r belongs to Rm*n, rmnRepresent user Um to article VnScoring；
The more attribute fusion similarity calculation Msim (U of step 7. useri,Vj)；The personal essential information of user, generally comprises： Gender, age, educational background, occupation, hobby etc.；User information is expressed as (X1, X2, X3, X4, X5, Xk), k represents user property Number；According to different attribute feature, suitable similarity calculating method is selected respectively；So, Msim (Ui,Vj)=a1sim (U1X1,V2X1)+a2sim(U1X2,V2X2)+…+aksim(U1Xk,V2Xk)；
Step 8. builds user's trusted neighborhood；Neighbours number max-thresholds P is set, is merged according to the more attributes of user Similarity calculation is as a result, obtain the neighbor user that target user is not more than P；
Step 9. builds project-item class table, and according to user items rating matrix, by mapping relations, it is more to obtain user Interest is drawn a portrait；
The more interest portraits of trusted neighborhood and step 9 user that step 10. is obtained with step 8 for constraint, use by filling Family-project rating matrix obtains non-sparse rating matrix；
Step 11. project high to the score value of target user does Top-N sequences, as recommendation results.
Described in this specification above content is only illustrations made for the present invention.Technology belonging to the present invention The technical staff in field can do described specific embodiment various modifications or additions or in a similar way It substitutes, content without departing from description of the invention or surmounts range defined in the claims, this should all be belonged to The protection domain of invention.

Claims (1)

  1. A kind of 1. recommendation method based on the more interest of user, it is characterised in that：Determining target user's trusted neighbor user mistake Cheng Zhong proposes neighbours' screening technique of the more attribute fusions of user；According to the data to have scored in user-project, structure user is more Interest is drawn a portrait；It proposes to draw a portrait based on more interest and the score in predicting method of similitude weighting is merged with more attributes；It is proposed a kind of user More interest quaternary group models, are as follows：
    Step 1. user is more, and interest model is expressed as：(R, I, C, P) four-tuple, R gather for consumer articles scoring, and I expressions user- Article hobby set, C represent article-category set, and P represents user-category set；
    Step 2.R gathers for consumer articles scoring, R=(R1,R2,…,Rn)T, n indicates n user, if 0<i<=n, RiIt represents User's score information of i-th of user, each user's score information table are made of many scorings to article, Ri=(Ri1, Ri2,…,Rim), m represents m article, R in matrix RijNumerical value represent user uiTo article vjScoring, note score value be Score, ordinary circumstance agreement 1<=score<=5,1 representative does not like, and 5 representatives enjoy a lot, and -1 representative is not scored；
    Step 3.I represents user-article hobby set, Iil(1<=i<=n, 1<=l<=j) numerical value represent fancy grade, number Value is bigger, and expression is more liked, and value represents not liking for 0 or 1,0, and 1 represents to like；
    Step 4.C represents article-category set, represents the relationships of article and goods categories, belonging to the value of element indicates whether Such, it is 1 to belong to value, and it is 0 to be not belonging to value；
    Step 5.P represents user-category set, reflects relationship namely user's portrait between user and goods categories M；
    Step 6. data prediction obtains user's essential attribute information and the user items matrix that scored；Remember user Ji Yonghuji Close U={ U1,U2,U3,U4,U5,Um, article set V={ V1,V2,V3…VnConsumer articles grade form matrix Rm*nIt represents, r Belong to Rm*n, rmnRepresent user Um to article VnScoring；
    The more attribute fusion similarity calculation Msim (U of step 7. useri,Vj)；The personal essential information of user, generally comprises：Property Not, age, educational background, occupation, hobby etc.；User information is expressed as (X1, X2, X3, X4, X5, Xk), k represents user property Number；According to different attribute feature, suitable similarity calculating method is selected respectively；So, Msim (Ui,Vj)=a1sim (U1X1, V2X1)+a2sim(U1X2,V2X2)+…+aksim(U1Xk,V2Xk)；
    Step 8. builds user's trusted neighborhood；Neighbours number max-thresholds P is set, is merged according to the more attributes of user similar Result of calculation is spent, obtains the neighbor user that target user is not more than P；
    Step 9. builds project-item class table, according to user items rating matrix, by mapping relations, obtains the more interest of user Portrait；
    The more interest portraits of trusted neighborhood and step 9 user that step 10. is obtained with step 8 fill user-item for constraint Mesh rating matrix obtains non-sparse rating matrix；
    Step 11. project high to the score value of target user does Top-N sequences, as recommendation results.
CN201611203999.5A 2016-12-23 2016-12-23 A kind of recommendation method based on the more interest of user Pending CN108241619A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201611203999.5A CN108241619A (en) 2016-12-23 2016-12-23 A kind of recommendation method based on the more interest of user

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201611203999.5A CN108241619A (en) 2016-12-23 2016-12-23 A kind of recommendation method based on the more interest of user

Publications (1)

Publication Number Publication Date
CN108241619A true CN108241619A (en) 2018-07-03

Family

ID=62703979

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201611203999.5A Pending CN108241619A (en) 2016-12-23 2016-12-23 A kind of recommendation method based on the more interest of user

Country Status (1)

Country Link
CN (1) CN108241619A (en)

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109214882A (en) * 2018-07-09 2019-01-15 西北大学 A kind of Method of Commodity Recommendation
CN109670107A (en) * 2018-12-07 2019-04-23 杭州飞弛网络科技有限公司 A kind of stranger's social activity recommended method and system based on interest big data
CN109829110A (en) * 2019-01-29 2019-05-31 四川长虹电器股份有限公司 A kind of personalized recommendation method of learning materials
CN111143699A (en) * 2020-01-03 2020-05-12 上海理工大学 Recommendation system based on similarity and confidence clustering
CN111241418A (en) * 2020-01-07 2020-06-05 北京邮电大学 Information recommendation method and device based on local weighted centrality trust inference
CN112883268A (en) * 2021-02-22 2021-06-01 中国计量大学 Session recommendation method considering user multiple interests and social influence

Cited By (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109214882A (en) * 2018-07-09 2019-01-15 西北大学 A kind of Method of Commodity Recommendation
CN109214882B (en) * 2018-07-09 2021-06-25 西北大学 Commodity recommendation method
CN109670107A (en) * 2018-12-07 2019-04-23 杭州飞弛网络科技有限公司 A kind of stranger's social activity recommended method and system based on interest big data
CN109670107B (en) * 2018-12-07 2020-10-16 杭州飞弛网络科技有限公司 Stranger social activity recommendation method and system based on big interest data
CN109829110A (en) * 2019-01-29 2019-05-31 四川长虹电器股份有限公司 A kind of personalized recommendation method of learning materials
CN111143699A (en) * 2020-01-03 2020-05-12 上海理工大学 Recommendation system based on similarity and confidence clustering
CN111143699B (en) * 2020-01-03 2023-07-28 上海理工大学 Recommendation system based on similarity and confidence coefficient clustering
CN111241418A (en) * 2020-01-07 2020-06-05 北京邮电大学 Information recommendation method and device based on local weighted centrality trust inference
CN111241418B (en) * 2020-01-07 2023-04-18 北京邮电大学 Information recommendation method and device based on local weighted centrality trust inference
CN112883268A (en) * 2021-02-22 2021-06-01 中国计量大学 Session recommendation method considering user multiple interests and social influence
CN112883268B (en) * 2021-02-22 2022-02-01 中国计量大学 Session recommendation method considering user multiple interests and social influence

Similar Documents

Publication Publication Date Title
CN108241619A (en) A kind of recommendation method based on the more interest of user
Dzyabura et al. Visual elicitation of brand perception
CN105740430B (en) A kind of personalized recommendation method of mosaic society&#39;s information
Qian et al. Personalized recommendation combining user interest and social circle
Assaker et al. Moderating effects of tourists’ novelty-seeking tendencies on destination image, visitor satisfaction, and short-and long-term revisit intentions
CN107038609A (en) A kind of Method of Commodity Recommendation and system based on deep learning
CN103246980B (en) Information output method and server
CN108920641A (en) A kind of information fusion personalized recommendation method
Feng et al. Recommendation via user's personality and social contextual
CN108334592B (en) Personalized recommendation method based on combination of content and collaborative filtering
CN107220365A (en) Accurate commending system and method based on collaborative filtering and correlation rule parallel processing
CN106959966A (en) A kind of information recommendation method and system
CN108228867A (en) A kind of theme collaborative filtering recommending method based on viewpoint enhancing
CN107833117A (en) A kind of Bayes&#39;s personalized ordering for considering label information recommends method
CN112507246B (en) Social recommendation method fusing global and local social interest influence
CN110083764A (en) A kind of collaborative filtering cold start-up way to solve the problem
KR101458588B1 (en) Expert curation recommendation system and expert recommendation method using thereof by field
CN108615177A (en) Electric terminal personalized recommendation method based on weighting extraction interest-degree
CN109885776A (en) Recommended models can be explained in open source community PR reviewer
Wang et al. Multifunctional product marketing using social media based on the variable-scale clustering
CN110390046B (en) Collaborative filtering recommendation method and system
CN115329215A (en) Recommendation method and system based on self-adaptive dynamic knowledge graph in heterogeneous network
CN104462597B (en) A kind of positive negativity of synthetic user scores and the collaborative filtering method of scoring preference heterogeneity
Cai et al. An extension of social network group decision-making based on trustrank and personas
CN110377841B (en) Similarity calculation method and system applied to collaborative filtering method

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
WD01 Invention patent application deemed withdrawn after publication
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20180703