CN108460619A - A kind of fusion shows the Collaborative Recommendation model of implicit feedback - Google Patents
A kind of fusion shows the Collaborative Recommendation model of implicit feedback Download PDFInfo
- Publication number
- CN108460619A CN108460619A CN201810038717.3A CN201810038717A CN108460619A CN 108460619 A CN108460619 A CN 108460619A CN 201810038717 A CN201810038717 A CN 201810038717A CN 108460619 A CN108460619 A CN 108460619A
- Authority
- CN
- China
- Prior art keywords
- user
- article
- scoring
- implicit feedback
- feedback data
- 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
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
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0241—Advertisements
- G06Q30/0251—Targeted advertisements
- G06Q30/0255—Targeted advertisements based on user history
-
- 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
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0241—Advertisements
- G06Q30/0251—Targeted advertisements
- G06Q30/0269—Targeted advertisements based on user profile or attribute
- G06Q30/0271—Personalized advertisement
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Accounting & Taxation (AREA)
- Development Economics (AREA)
- Finance (AREA)
- Strategic Management (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Databases & Information Systems (AREA)
- General Physics & Mathematics (AREA)
- Entrepreneurship & Innovation (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Economics (AREA)
- Game Theory and Decision Science (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses the Collaborative Recommendation models that a kind of fusion shows implicit feedback, from the angle of the true shopping process of analog subscriber, our recommendation task is completed by two steps, after being modeled first by the user concealed feedback data of the recommended models of sequencing-oriented analysis, the article selected user and most possibly browsed is concentrated from article to be sorted, then pass through user's explicit feedback data using the recommended models towards scoring, learn user characteristics matrix and article characteristics matrix for predicting scoring, it is then based on scoring and sequencing is rearranged to the article collection selected before, target user is returned to as last recommendation list, to greatly improve the accuracy of recommendation, it is worth with extensive commercial application.
Description
Technical field
The invention belongs to computer software technical field, specifically a kind of fusion shows the Collaborative Recommendation model of implicit feedback.
Background technology
In current most of online shopping platforms, the historical behavior data (such as score, thumb up, browsing record) of user
Implicitly include the preference information of user, many commending systems are all as one of important sources of input data.According to
The online behavioral data of user, can be divided into explicit feedback data and implicit feedback data by the feedback mechanism at family, and for
Current electric business field, the two data are and deposit, and are seldom individually present explicit feedback or implicit feedback data.However,
Center of gravity is all concentrated on individual explicit feedback or implicit feedback by most of personalized recommendation technologies before, it is difficult to cope with current
Active demand of the electric business for recommended technology, how to merge aobvious implicit feedback for recommend task be always study difficult point it
One.In addition, the sequencing inside the recommendation article collection finally generated can influence the buying intention of people, generally believe forward
Article is that user most thinks purchase, therefore, it is recommended that task should also consider the sequencing between different articles in recommendation results.
In the present invention, it is believed that process of purchase of the user under reality scene is such：When user first passes through long
Between browse commodity, finally can just pick out the commodity for oneself wanting to buy.User can leave many implicit feedback data in navigation process
(such as click, browse, purchaser record), and evaluation information after point of purchase is explicit feedback data (such as scoring, grading of user
Deng), so final recommendation results should not only depend on explicit feedback data or implicit feedback data, and recommend to appoint
Business can also be divided into two steps：The article collection that user most possibly browses, then the rearrangement of the article collection to obtaining first are generated, is generated
User most possibly buys article collection.
Invention content
The purpose of the present invention is to provide the Collaborative Recommendation models that a kind of fusion shows implicit feedback.The model recommendation effect
It is good, and can realize personalized recommendation.
The technical solution adopted by the present invention to solve the technical problems includes the following steps：
Step 1, the recommended models that required sequencing-oriented is trained using implicit feedback data, and selecting that user most has can
The article collection I that can be browsed；
Step 2 is added explicit feedback data and implicit feedback data in the model towards scoring and carrys out training pattern, study
Go out user characteristics matrix and article characteristics matrix to score for predicting；It resequences, obtains to the article collection I that step 1 obtains
The article collection II bought to user's most probable, target user is returned to by article collection II as final recommendation results；
By above step, the present invention pushes away the implicit feedback data and explicit feedback data of user for personalization together
It recommends, realizes the purpose for improving personalized recommendation effect.
Step 1 is implemented as follows：
The proposed algorithm that the recommended models of sequencing-oriented use in step 1 may be used at present to grade Ranking Algorithm
In be applied to implicit feedback data in a kind of more outstanding algorithm BPR (Bayesian Personalized Ranking),
The realization of the algorithm：
First, the pretreatment of data pairization is carried out to the history score data of user：BPR algorithms comment article user
The set that divided data (user had interactive article to be denoted as " 1 ", and the article not interacted is denoted as " 0 ") processing is one pair pairs
<I, j>, wherein i is the article that scoring is 1, and j is the article that scoring is 0.Use triple<U, i, j>To express：User " u " likes
Article " i " is more than article " j ".
The precondition of the recommended models：
1. the preference behavior between user is respectively independent.
2. same user is respectively independent to the partial ordering relation of different articles.
Therefore, definition needs the target to maximize：
Π p (Θ | i ＞uJ) ∝ Π p (i ＞uj|Θ)p(Θ)(1)
Wherein Θ is required recommended models parameter, including：The eigenmatrix U of the user and eigenmatrix V of article.p(Θ
| i ＞uJ) posterior probability, p (i ＞ are indicateduJ | Θ) indicate that likelihood part, p (Θ) indicate prior probability, i ＞uJ indicates user " u "
Like article " i " and is more than article " j ".
Wherein about likelihood part：P (i ＞uJ | Θ)=δ (xU, i-xU, j),xU, i=pu·piIf
Prior probability obeys following distribution：Θ~N (0, λΘI), then the density function of prior probability is：
Based on above it is assumed that optimization aim further spreads out to obtain：
Π p (i ＞uJ | Θ) p (Θ) ∝ ln (p (i ＞uJ | Θ) p (Θ))=∑ (ln δ (xU, i-xU, j)+lnp(Θ))
=∑ (ln δ (xU, i-xU, j)-λΘ||Θ||2)
=∑ (ln δ (xU, i-xU, j)-λΘ||pu||2-λΘ||qi||2-λΘ||qj||2)
=∑ (ln δ (pu·qi-pu·qj)-λΘ||pu||2-λΘ||qi||2-λΘ||qj||2)(3)
Wherein xU, i<pu, qi>Preferences of the user u to project i is indicated for two vectorial inner product forms, and Θ is to push away
Recommend model parameter.In order to maximize the functional value of expression formula (3) above, solved by stochastic gradient descent algorithm (SGD).This
Invention can obtain the article collection I that user most possibly browses by the method.
Step 2 is implemented as follows：
For the recommended models towards scoring in step 2, can be purchased using SVD++ models to complete screening user's most probable
The article collection II bought.SVD++ models be on the basis of singular value decomposition model (SVD) by incorporate implicit feedback data to
Realize improvement.The core of SVD is matrix decomposition technology, basic thought be exactly according to the historical behavior data set of user (such as
The score data of user) some features are extracted, as the basis for recommending task.For film recommendation, these features can
With the degree of making laughs, terrified degree, degree of risk etc. for being interpreted as film.
In SVD models, there is the articles not scored by each user, and for such article, actually user also may be used
Can there is potential interest, and SVD is mistakenly considered user and loses interest in this.So having incorporated use in SVD++ models
The fact that there are potential interest at family.
Wherein N (u) represents user's u scoring article collection, and W is called potential article characteristics matrix.Objective function is as follows：
Wherein,To
Measure biIndicate deviation of the scoring of film i relative to average score, vectorial buIndicate the scorings made of user u relative to averagely commenting
The deviation divided, μ is denoted as by average score.
The present invention has the beneficial effect that：
The present invention is a kind of Collaborative Recommendation model of the aobvious implicit feedback of fusion.In traditional recommended technology, towards scoring
Recommended models are with only user's explicit feedback data, to lose the ability for liking similarity between capture user, and
The recommended models of sequencing-oriented then only consider whether user is consistent to the preference of product pair, and ignore the preference between user
This important information of degree.The present invention analyzes and has studied two kinds of traditional recommended models, it was found that respective shortcoming.Cause
This, the present invention ties the advantages of both recommended models from the angle of the true shopping process of analog subscriber, by two steps
Be combined together the recommendation task to complete us, first by the recommended models of sequencing-oriented analyze user concealed feedback data come
After modeling, the article selected user and most possibly browsed is concentrated from article to be sorted, then uses the recommended models towards scoring
By user's explicit feedback data, learns user characteristics matrix and article characteristics matrix for predicting scoring, be then based on and comment
Divide and sequencing is rearranged to the article collection selected before, target user is returned to as last recommendation list.
Description of the drawings
Fig. 1 is flow chart of the present invention.
Specific implementation mode
The invention will be further described with reference to the accompanying drawings and examples.
A kind of fusion shows the Collaborative Recommendation model of implicit feedback, and general frame is as shown in Figure 1.Include the following steps：
Step 1, the recommended models (such as BPR) that required sequencing-oriented is trained using implicit feedback data, and select use
The article collection I that family most possibly browses.
Explicit feedback data and implicit feedback data are added in the model (such as SVD++) towards scoring to train mould in step 2
Type resequences to the article collection I that step 1 obtains, and obtains the article collection II of user's most probable purchase, article collection II is made
For final recommendation results.
By above step, the present invention pushes away the implicit feedback data and explicit feedback data of user for personalization together
It recommends, realizes the purpose for improving personalized recommendation effect.
Step 1 is implemented as follows：
The proposed algorithm that the recommended models of sequencing-oriented use in step 1 may be used at present to grade Ranking Algorithm
In be applied to implicit feedback data in a kind of more outstanding algorithm BPR (Bayesian Personalized Ranking),
The realization of the algorithm：
First, the pretreatment of data pairization is carried out to the history score data of user：BPR algorithms comment article user
The set that divided data (user had interactive article to be denoted as " 1 ", and the article not interacted is denoted as " 0 ") processing is one pair pairs
<I, j>, wherein i is the article that scoring is 1, and j is the article that scoring is 0.Use triple<U, i, j>To express：User " u " likes
Article " i " is more than article " j ".
The precondition of the recommended models：
1. the preference behavior between user is respectively independent.
2. same user is respectively independent to the partial ordering relation of different articles.
Therefore, definition needs the target to maximize：
Π p (Θ | i ＞uJ) ∝ Π p (i ＞uj|Θ)p(Θ)(1)
Wherein Θ is required recommended models parameter, including：The eigenmatrix U of the user and eigenmatrix V of article.p(Θ
| i ＞uJ) posterior probability, p (i ＞ are indicateduJ | Θ) indicate that likelihood part, p (Θ) indicate prior probability, i ＞uJ indicates user " u "
Like article " i " and is more than article " j ".
Wherein about likelihood part：P (i ＞uJ | Θ)=δ (xU, i-xU, j),xU, i=pu·piIf
Prior probability obeys following distribution：Θ~N (O, λθI), then the density function of prior probability is：
Based on above it is assumed that optimization aim further spreads out to obtain：
Π p (i ＞uJ | Θ) p (Θ) ∝ ln (p (i ＞uJ | Θ) p (Θ))=∑ (ln δ (xU, i, xU, j)+ln p(Θ))
=∑ (ln δ (xU, i-xU, j)-λΘ||Θ||2)
=∑ (ln δ (xU, i-xU, j)-λΘ||pu||2-λΘ||qi||2-λΘ||qj||2)
=∑ (ln δ (pu·qi-pu·qj)-λΘ||pu||2-λΘ||qi||2-λΘ||qj||2)(3)
Wherein xU, i<pu, qi>Preferences of the user u to project i is indicated for two vectorial inner product forms, and Θ is to push away
Recommend model parameter.In order to maximize the functional value of expression formula (3) above, solved by stochastic gradient descent algorithm (SGD).This
Invention can obtain the article collection I that user most possibly browses by the method.
Step 2 is implemented as follows：
For the recommended models towards scoring in step 2, can be purchased using SVD++ models to complete screening user's most probable
The article collection II bought.SVD++ models be on the basis of singular value decomposition model (SVD) by incorporate implicit feedback data to
Realize improvement.The core of SVD is matrix decomposition technology, basic thought be exactly according to the historical behavior data set of user (such as
The score data of user) some features are extracted, as the basis for recommending task.For film recommendation, these features can
With the degree of making laughs, terrified degree, degree of risk etc. for being interpreted as film.
In SVD models, there is the articles not scored by each user, and for such article, actually user also may be used
Can there is potential interest, and SVD is mistakenly considered user and loses interest in this.So having incorporated use in SVD++ models
The fact that there are potential interest at family.
Wherein N (u) represents user's u scoring article collection, and W is called potential article characteristics matrix.Objective function is as follows：
Wherein,To
Measure biIndicate deviation of the scoring of film i relative to average score, vectorial buIndicate the scorings made of user u relative to averagely commenting
The deviation divided, μ is denoted as by average score.
Claims (2)
1. a kind of fusion shows the Collaborative Recommendation model of implicit feedback, it is characterised in that from the angle of the true shopping process of analog subscriber
Degree sets out, and is as follows：
Step 1, the recommended models that required sequencing-oriented is trained using implicit feedback data, and it is most possibly clear to select user
The article collection I look at；
Step 2 is added explicit feedback data and implicit feedback data in the model towards scoring and carrys out training pattern, learn use
Family eigenmatrix and article characteristics matrix score for predicting；It resequences, is used to the article collection I that step 1 obtains
The article collection II of family most probable purchase, target user is returned to using article collection II as final recommendation results；
The proposed algorithm that the recommended models of sequencing-oriented use in step 1, using at present to being applied in grade Ranking Algorithm
BPR algorithms in implicit feedback data, are implemented as follows：
First, the pretreatment of data pairization is carried out to the history score data of user：BPR algorithms are by user to the scoring number of article
Set ＜ i, the j ＞, wherein i for being one pair couples according to processing are the article that scoring is 1, and j is the article that scoring is 0；Use ternary
＜ u, i, j ＞ are organized to express：User " u " likes article " i " and is more than article " j "；User had interactive article to be denoted as " 1 ", did not had
There is interactive article to be denoted as " 0 "；
The precondition of the recommended models of sequencing-oriented：
1. the preference behavior between user is respectively independent；
2. same user is respectively independent to the partial ordering relation of different articles；
Therefore, definition needs the target to maximize：
Π p (Θ | i ＞uJ) ∝ Π p (i ＞uj|Θ)p(Θ) (1)
Wherein Θ is required recommended models parameter, including：The eigenmatrix U of the user and eigenmatrix V of article；P (Θ | i ＞uJ) posterior probability, p (i ＞ are indicateduJ | Θ) indicate that likelihood part, p (Θ) indicate prior probability, i ＞uJ indicates that user " u " likes
Article " i " is more than article " j "；
Wherein about likelihood part：P (i ＞uJ | Θ)=δ (xU, i-xU, j),xU, i=pu·piIf priori
Probability obeys following distribution：Θ~N (0, λΘI), then the density function of prior probability is：
Based on above it is assumed that optimization aim further spreads out to obtain：
Wherein xU, i＜ pu, qi＞ indicates preferences of the user u to project i for two vectorial inner product forms, and Θ is to recommend
Model parameter；In order to maximize the functional value of expression formula (3) above, is solved by stochastic gradient descent algorithm, obtain user
The article collection I most possibly browsed.
2. a kind of fusion according to claim 1 shows the Collaborative Recommendation model of implicit feedback, it is characterised in that for step 2
In the recommended models towards scoring, using SVD++ models come complete screening user's most probable purchase article collection II, SVD++ mould
The fact that incorporated user in type there are potential interest；
Wherein N (u) represents user's u scoring article collection, and W is called potential article characteristics matrix；Objective function is as follows：
Wherein,Vectorial biTable
Show deviation of the scoring relative to average score of film i, vectorial buThe scoring that expression user u makes is relative to the inclined of average score
Average score is denoted as μ by difference.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810038717.3A CN108460619B (en) | 2018-01-16 | 2018-01-16 | Method for providing collaborative recommendation model fusing explicit and implicit feedback |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810038717.3A CN108460619B (en) | 2018-01-16 | 2018-01-16 | Method for providing collaborative recommendation model fusing explicit and implicit feedback |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108460619A true CN108460619A (en) | 2018-08-28 |
| CN108460619B CN108460619B (en) | 2020-04-21 |
Family
ID=63220184
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810038717.3A Active CN108460619B (en) | 2018-01-16 | 2018-01-16 | Method for providing collaborative recommendation model fusing explicit and implicit feedback |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108460619B (en) |
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109299370A (en) * | 2018-10-09 | 2019-02-01 | 中国科学技术大学 | Multipair grade personalized recommendation method |
| CN109446430A (en) * | 2018-11-29 | 2019-03-08 | 西安电子科技大学 | Method, apparatus, computer equipment and the readable storage medium storing program for executing of Products Show |
| CN109885644A (en) * | 2019-04-08 | 2019-06-14 | 浙江大学城市学院 | A kind of importance appraisal procedure for Internet of Things Item Information searching order |
| CN110020207A (en) * | 2019-04-16 | 2019-07-16 | 中森云链(成都)科技有限责任公司 | A kind of interpretable Top-K recommended method of examination question of fusion aspect implicit feedback |
| CN110059251A (en) * | 2019-04-22 | 2019-07-26 | 郑州大学 | Collaborative filtering recommending method based on more relationship implicit feedback confidence levels |
| CN110110139A (en) * | 2019-04-19 | 2019-08-09 | 北京奇艺世纪科技有限公司 | The method, apparatus and electronic equipment that a kind of pair of recommendation results explain |
| CN111104601A (en) * | 2019-12-26 | 2020-05-05 | 河南理工大学 | Antagonistic multi-feedback-level paired personalized ranking method |
| CN111160859A (en) * | 2019-12-26 | 2020-05-15 | 淮阴工学院 | Human resource post recommendation method based on SVD + + and collaborative filtering |
| CN111241415A (en) * | 2019-12-28 | 2020-06-05 | 四川文理学院 | Recommendation method fusing multi-factor social activity |
| CN111310029A (en) * | 2020-01-20 | 2020-06-19 | 哈尔滨理工大学 | Mixed recommendation method based on user commodity portrait and potential factor feature extraction |
| CN113010802A (en) * | 2021-03-25 | 2021-06-22 | 华南理工大学 | Recommendation method facing implicit feedback based on multi-attribute interaction of user and article |
| CN113111251A (en) * | 2020-01-10 | 2021-07-13 | 阿里巴巴集团控股有限公司 | Project recommendation method, device and system |
| US20210241347A1 (en) * | 2020-01-31 | 2021-08-05 | Walmart Apollo, Llc | Single-select predictive platform model |
| CN113538093A (en) * | 2021-07-16 | 2021-10-22 | 四川大学 | High-quality power package recommendation method based on improved collaborative filtering algorithm |
| US11861676B2 (en) | 2020-01-31 | 2024-01-02 | Walmart Apollo, Llc | Automatic item grouping and personalized department layout for reorder recommendations |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104281956A (en) * | 2014-10-27 | 2015-01-14 | 南京信息工程大学 | Dynamic recommendation method capable of adapting to user interest changes based on time information |
| US20150187024A1 (en) * | 2013-12-27 | 2015-07-02 | Telefonica Digital España, S.L.U. | System and Method for Socially Aware Recommendations Based on Implicit User Feedback |
| CN107423335A (en) * | 2017-04-27 | 2017-12-01 | 电子科技大学 | A kind of negative sample system of selection for single class collaborative filtering problem |
-
2018
- 2018-01-16 CN CN201810038717.3A patent/CN108460619B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150187024A1 (en) * | 2013-12-27 | 2015-07-02 | Telefonica Digital España, S.L.U. | System and Method for Socially Aware Recommendations Based on Implicit User Feedback |
| CN104281956A (en) * | 2014-10-27 | 2015-01-14 | 南京信息工程大学 | Dynamic recommendation method capable of adapting to user interest changes based on time information |
| CN107423335A (en) * | 2017-04-27 | 2017-12-01 | 电子科技大学 | A kind of negative sample system of selection for single class collaborative filtering problem |
Cited By (23)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109299370A (en) * | 2018-10-09 | 2019-02-01 | 中国科学技术大学 | Multipair grade personalized recommendation method |
| CN109299370B (en) * | 2018-10-09 | 2022-03-01 | 中国科学技术大学 | Multi-pair level personalized recommendation method |
| CN109446430A (en) * | 2018-11-29 | 2019-03-08 | 西安电子科技大学 | Method, apparatus, computer equipment and the readable storage medium storing program for executing of Products Show |
| CN109885644B (en) * | 2019-04-08 | 2021-04-06 | 浙江大学城市学院 | Importance evaluation method for searching and sorting of Internet of things item information |
| CN109885644A (en) * | 2019-04-08 | 2019-06-14 | 浙江大学城市学院 | A kind of importance appraisal procedure for Internet of Things Item Information searching order |
| CN110020207A (en) * | 2019-04-16 | 2019-07-16 | 中森云链(成都)科技有限责任公司 | A kind of interpretable Top-K recommended method of examination question of fusion aspect implicit feedback |
| CN110110139A (en) * | 2019-04-19 | 2019-08-09 | 北京奇艺世纪科技有限公司 | The method, apparatus and electronic equipment that a kind of pair of recommendation results explain |
| CN110059251A (en) * | 2019-04-22 | 2019-07-26 | 郑州大学 | Collaborative filtering recommending method based on more relationship implicit feedback confidence levels |
| CN110059251B (en) * | 2019-04-22 | 2022-10-28 | 郑州大学 | Collaborative filtering recommendation method based on multi-relation implicit feedback confidence |
| CN111104601A (en) * | 2019-12-26 | 2020-05-05 | 河南理工大学 | Antagonistic multi-feedback-level paired personalized ranking method |
| CN111160859A (en) * | 2019-12-26 | 2020-05-15 | 淮阴工学院 | Human resource post recommendation method based on SVD + + and collaborative filtering |
| CN111104601B (en) * | 2019-12-26 | 2022-09-13 | 河南理工大学 | Antagonistic multi-feedback-level paired personalized ranking method |
| CN111241415A (en) * | 2019-12-28 | 2020-06-05 | 四川文理学院 | Recommendation method fusing multi-factor social activity |
| CN111241415B (en) * | 2019-12-28 | 2023-07-21 | 四川文理学院 | Recommendation method integrating multi-factor social activities |
| CN113111251A (en) * | 2020-01-10 | 2021-07-13 | 阿里巴巴集团控股有限公司 | Project recommendation method, device and system |
| CN111310029A (en) * | 2020-01-20 | 2020-06-19 | 哈尔滨理工大学 | Mixed recommendation method based on user commodity portrait and potential factor feature extraction |
| CN111310029B (en) * | 2020-01-20 | 2022-11-01 | 哈尔滨理工大学 | Mixed recommendation method based on user commodity portrait and potential factor feature extraction |
| US20210241347A1 (en) * | 2020-01-31 | 2021-08-05 | Walmart Apollo, Llc | Single-select predictive platform model |
| US11636525B2 (en) * | 2020-01-31 | 2023-04-25 | Walmart Apollo, Llc | Single-select predictive platform model |
| US11861676B2 (en) | 2020-01-31 | 2024-01-02 | Walmart Apollo, Llc | Automatic item grouping and personalized department layout for reorder recommendations |
| CN113010802A (en) * | 2021-03-25 | 2021-06-22 | 华南理工大学 | Recommendation method facing implicit feedback based on multi-attribute interaction of user and article |
| CN113010802B (en) * | 2021-03-25 | 2022-09-20 | 华南理工大学 | Recommendation method facing implicit feedback based on multi-attribute interaction of user and article |
| CN113538093A (en) * | 2021-07-16 | 2021-10-22 | 四川大学 | High-quality power package recommendation method based on improved collaborative filtering algorithm |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108460619B (en) | 2020-04-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108460619A (en) | A kind of fusion shows the Collaborative Recommendation model of implicit feedback | |
| CN106202519A (en) | A kind of combination user comment content and the item recommendation method of scoring | |
| CN109087178A (en) | Method of Commodity Recommendation and device | |
| CN111709810A (en) | Object recommendation method and device based on recommendation model | |
| CN109102127A (en) | Method of Commodity Recommendation and device | |
| CN111681084B (en) | E-commerce platform recommendation method based on social relationship influence factors | |
| CN109190030A (en) | Merge the implicit feedback recommended method of node2vec and deep neural network | |
| CN109543840A (en) | A kind of Dynamic recommendation design method based on multidimensional classification intensified learning | |
| CN105787100A (en) | User session recommendation method based on deep neural network | |
| CN108109058B (en) | Single-classification collaborative filtering method integrating personality traits and article labels | |
| CN108829847A (en) | Commodity search method and system based on multi-modal shopping preferences | |
| CN111949887A (en) | Item recommendation method and device and computer-readable storage medium | |
| CN107103093B (en) | Short text recommendation method and device based on user behavior and emotion analysis | |
| CN113592609A (en) | Personalized clothing matching recommendation method and system using time factors | |
| Luo et al. | Conceptualize and infer user needs in e-commerce | |
| Ahamed et al. | A recommender system based on deep neural network and matrix factorization for collaborative filtering | |
| CN109684561B (en) | Interest point recommendation method based on deep semantic analysis of user sign-in behavior change | |
| CN113190751B (en) | Recommendation method fusing keyword generation | |
| CN105260458A (en) | Video recommendation method for display apparatus and display apparatus | |
| Vartak et al. | CHIC: a combination-based recommendation system | |
| CN111310029B (en) | Mixed recommendation method based on user commodity portrait and potential factor feature extraction | |
| CN111159556B (en) | Bayesian personalized ordering recommendation method based on curiosity of user | |
| CN110020918B (en) | Recommendation information generation method and system | |
| CN111815410A (en) | Commodity recommendation method based on selective neighborhood information | |
| CN111738787A (en) | Information pushing method and device |
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 | ||
| EE01 | Entry into force of recordation of patent licensing contract | ||
| EE01 | Entry into force of recordation of patent licensing contract |
Application publication date: 20180828 Assignee: Wenzhou Kaichen Technology Co.,Ltd. Assignor: HANGZHOU DIANZI University Contract record no.: X2021330000778 Denomination of invention: A method of providing collaborative recommendation model integrating explicit and implicit feedback Granted publication date: 20200421 License type: Common License Record date: 20211130 |
|
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20221206 Address after: 325036 Room 301, Floor 3, Building 10, Incubator of Wenzhou National University Science Park, No. 38, Dongfang South Road, Ouhai Economic Development Zone, Wenzhou City, Zhejiang Province Patentee after: Wenzhou Kaichen Technology Co.,Ltd. Address before: 310018 No. 2 street, Xiasha Higher Education Zone, Hangzhou, Zhejiang Patentee before: HANGZHOU DIANZI University |