CN108549729A - Personalized user collaborative filtering recommending method based on Covering reduct - Google Patents
Personalized user collaborative filtering recommending method based on Covering reduct Download PDFInfo
- Publication number
- CN108549729A CN108549729A CN201810486715.0A CN201810486715A CN108549729A CN 108549729 A CN108549729 A CN 108549729A CN 201810486715 A CN201810486715 A CN 201810486715A CN 108549729 A CN108549729 A CN 108549729A
- Authority
- CN
- China
- Prior art keywords
- user
- article
- target user
- indicate
- covering
- 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
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0631—Item recommendations
Abstract
The present invention discloses a kind of personalized user collaborative filtering recommending method based on Covering reduct, exactly define the concept of the redundant subscribers of target user, the function of redundant elements can be removed according to Covering reduct in covering rough set, the redundant subscribers of target user are removed, to ensure that target user adjacent user quality, to using these high quality adjacent user score information be embodied as target user provide high-precision and diversified personalized recommendation.
Description
Technical field
Recommendation accuracy can be improved the present invention relates to commending system field more particularly to one kind and is based on multifarious
The personalized user collaborative filtering recommending method of Covering reduct.
Background technology
Commending system can intelligently perceive the interest or demand of user by the personal information of user, realize the height of information
Quality is recommended, and is efficiently solved the problems, such as " information overload ".User collaborative filter algorithm be commending system field it is most widely used,
One of most successful technology, if assuming that user is having similar hobby in the past, they may also have similar in future
Hobby, have many advantages, such as to calculate simple, efficiency and precision be high.But in existing user collaborative filter algorithm, target user
Adjacent user tend to possess identical hobby, so often being collected by score high article of the prediction that these adjacent users obtain
In in the article of a small amount of type, or even be only popular article, therefore its diversity recommended is often unsatisfactory.
Invention content
The present invention is to provide one kind in order to solve the above-mentioned technical problem present in the prior art and recommendation accuracy can be improved
And there is the multifarious personalized user collaborative filtering recommending method based on Covering reduct.
Technical solution of the invention is：A kind of personalized user collaborative filtering recommending method based on Covering reduct,
It is characterized in that carrying out in accordance with the following steps successively：
Step 1. statistics forms two-dimentional score information table：
Two-dimentional score information table is formed to the score information of article according to userRM={ U, I, R ∪ { * } }；The two dimension scoring letter
Cease tableRMIn,UIndicate the set of user,IIndicate the set of article,R∪{*}Indicate scoring set of the user to article, wherein*
Indicate that user does not score to article；
Enable useru∈UTo articlei∈IScoring ber u,i ∈R∪{*}, and useruAverage score be；θIt is commented for user
The threshold value divided, ifr u,i ≥θ, show useruLike articlei；UseruThe article collection to have scored is combined intoI u ={i∈I|r u,i ≠*}；For useruThe article set not scored；Goods attribute matrix isAM；In user's set U, liked if there is user a
The article set of love is contained in the favorite article set of user b, then user a is known as the redundant subscribers of target user；
Step 2. carries out yojan using Covering reduct algorithm to redundant subscribers:
Step 2.1 enables article setIAs domainI, in domainIIn, the article that each user likes forms a set；In object
Product attribute matrixAMMiddle extraction target user's likes attribute：
(1)
In formula (1),mIndicate the number of attribute,at m Indicate an attribute,av m Indicate attributeat m Value；
Step 2.2. likes attribute using the target user of acquisition, builds the decision set of target userD, decision setDBy having
The article collection of attribute is liked to be combined into：
(2)
In formula (2),at m (i)= av m Indicate articleiIn attributeat m On value beav m ；
Step 2.3. is by domainIThe decision set of target user is reduced to by article setD, i.e. domainD；For each useru ∈U, build useruIn domainDOn like article setC u ：
(3)
It enablesC*=D-∪C u , C={ C 1 ,C 2 …C n , C*} Target user is constituted in domainDOne coveringC；
Step 2.4 utilizes Covering reduct algorithm, by redundant elements from coveringCMiddle yojan obtains the covering after yojanreduct(C) and yojan after userU r :
(4)
Step 3. utilizes the user after yojanU r Build target userauCandidate adjacent useru；
Step 4. calculates the similarity of target user and candidate adjacent user, chooses the adjacent user of target user：
Target user is calculated using Pearson's similarity metric function (5)auWith candidate adjacent useru∈U r Between similarity,
(5)
In formula (5),sim(au,u)Indicate target userauWith candidate adjacent useru∈U r Between similarity,I au ={i∈I| r au,i ≠*}Indicate target userauThe article set evaluated,Indicate the average score value of target user；
Then before selection similarity is highKAdjacent user of a candidate adjacent user as target userN au (k)；
Step 5. article that do not score target user carries out prediction scoring：
According to the adjacent user of target userN au (k)Score information, using adjustment weighted sum function (6) to target userau
The article set not scoredPrediction scoring is carried out, the prediction grade form of target user is obtained；
(6)
In formula (6),P au,i Indicate target userauTo articleiPrediction scoring,U i ={u∈U|r u,i ≠*}Object was evaluated in expression
ProductiUser set；λAs a regularization factors：
(7)
Before step 6. selection prediction scoring is highNA article is as recommendation results.
The present invention exactly defines the concept of the redundant subscribers of target user, can be with according to Covering reduct in covering rough set
The function of removing redundant elements removes the redundant subscribers of target user, to ensure that target user adjacent user matter
Amount provides high-precision and diversified to be embodied as target user using the score information of the adjacent user of these high quality
Propertyization is recommended.
Description of the drawings
Fig. 1 is the flow diagram of the embodiment of the present invention；
Fig. 2 is the embodiment of the present invention and comparative example precision metrology（MAE and RMSE）With adjacent user's quantity of target user
Variation and corresponding result schematic diagram.
Fig. 3 is that the embodiment of the present invention is measured with comparative example diversity（Coverage）With adjacent user's number of target user
The variation of amount and corresponding result schematic diagram.
Specific implementation mode
A kind of personalized user collaborative filtering recommending method based on Covering reduct of the present invention, as shown in Figure 1 successively according to such as
Lower step carries out：
Step 1. statistics forms two-dimentional score information table：
Two-dimentional score information table is formed to the score information of article according to userRM={ U, I, R ∪ { * } }；The two dimension scoring letter
Cease tableRMIn,UIndicate the set of user,IIndicate the set of article,R∪{*}Indicate scoring set of the user to article, wherein*
Indicate that user does not score to article；
Enable useru∈UTo articlei∈IScoring ber u,i ∈R∪{*}, and useruAverage score be；θIt is commented for user
The threshold value divided, ifr u,i ≥θ, show useruLike articlei；UseruThe article collection to have scored is combined intoI u ={i∈I|r u,i ≠*}；For useruThe article set not scored；Goods attribute matrix isAM；In user's set U, liked if there is user a
The article set of love is contained in the favorite article set of user b, then user a is known as the redundant subscribers of target user；
Such as：User gathersU = { user 1, and user 2, and user 3, target user }, article setI ={ article 1, article 2, article
3, article 4, article 5, article 6 }, the value range for the R that scores is [1,5].Then two-dimentional score information table RM is as shown in table 1：
Table 1
Threshold value of user's scoring etc. and 3 are enabled, like article of the article for being more than or equal to 3 that scores as user, as shown in Table 1：
User's 1 likes that article is { article 2, article 4, article 6 }；
User's 2 likes that article is { article 4, article 6 }；
User's 3 likes that article is { article 2, article 3, article 6 }；
Target user's likes that article is { article 1, article 3, article 4 }；
Step 2. carries out yojan using Covering reduct algorithm to redundant subscribers:
Step 2.1 enables article setIAs domainI, in domainIIn, the article that each user likes forms a set；In object
Product attribute matrixAMMiddle extraction target user's likes attribute：
(1)
In formula (1),mIndicate the number of attribute,at m Indicate an attribute,av m Indicate attributeat m Value；
Such as enable article setI ={ article 1, article 2, article 3, article 4, article 5, article 6 } is used as domain, table 2 to indicate article
Attribute matrix AM, article set is liked according to table 2 and target user, statistics obtains the liking corresponding to article of target user
Attribute value：
Comedy=3, terrible=2, action=1, drama=1, music=1,
Like attribute of maximum two attributes of statistical value as target user is selected, then the attribute of liking of target user is：
[comedy=1] ∧ [terrible=1] ∧ [action=0] ∧ [drama=0] ∧ [music=0]
Table 2
| Comedy | It is terrible | Action | Drama | Music | |
| Article 1 | 1 | 0 | 1 | 1 | 0 |
| Article 2 | 1 | 1 | 0 | 1 | 0 |
| Article 3 | 1 | 1 | 0 | 0 | 0 |
| Article 4 | 1 | 1 | 0 | 0 | 1 |
| Article 5 | 0 | 0 | 1 | 1 | 0 |
| Article 6 | 1 | 1 | 1 | 0 | 1 |
Step 2.2. likes attribute using the target user of acquisition, builds the decision set of target userD, decision setDBy having
The article collection of attribute is liked to be combined into：
(2)
In formula (2),at m (i)= av m Indicate articleiIn attributeat m On value beav m ；
Such as like attribute with target user：[comedy=1] ∧ [terrible=1] ∧ [action=0] ∧ [drama=0] ∧ [music=0],
Build the decision set of target userD, which is possessed by all（Comedy, it is terrible）The article of attribute is constituted, i.e.,：
Decision set D={ article 2, article 3, article 4, article 6 } can be obtained according to table 2；
Step 2.3. in order to eliminate the redundant subscribers of target user to the maximum extent, by domainITarget is reduced to by article set
The decision set of userD, i.e. domainD；For each useru∈U, build useruIn domainDOn like article setC u ：
(3)
It enablesC*=D-∪C u , C={ C 1 ,C 2 …C n , C*} Target user is constituted in domainDOne coveringC；
Such asC 1 ={ article 2, article 4, article 6 }；
C 2 ={ article 4, article 6 }；
C 3 ={ article 2, article 3, article 6 }；
ThenC = { C 1 , C 2 , C 3 } With regard to constituting target user's decision setDOn one coveringC；
Step 2.4 utilizes Covering reduct algorithm, by redundant elements from coveringCMiddle yojan obtains the covering after yojanreduct(C)；Redundant elements yojan, which finishes, means that the redundant subscribers of target user are all deleted, to the use after yojan
FamilyU r :
(4)
Due toC 2 ⊂ C 1 , according to Covering reduct algorithm,C 2 It is referred to as redundant elements from coveringCMiddle removal, thereforereduct(C)=
{C 1 , C 3 }；The redundant subscribers that user 2 is thus referred to as target user are removed the user so after yojanU r ={ user 1, user
3}；
Step 3. utilizes the user after yojanU r Build target userauCandidate adjacent useru, i.e., target user's is candidate adjacent
Nearly user is { user 1, and user 3 }；
Step 4. calculates the similarity of target user and candidate adjacent user, chooses the adjacent user of target user：
Target user is calculated using Pearson's similarity metric function (5)auWith candidate adjacent useru∈U r Between similarity,
(5)
In formula (5),sim(au,u)Indicate target userauWith candidate adjacent useru∈U r Between similarity,I au ={i∈I| r au,i ≠*}Indicate target userauThe article set evaluated,Indicate the average score value of target user；
Then before selection similarity is highKAdjacent user of a candidate adjacent user as target userN au (k)；
Target user and user 1, the similarity of target user and user 3 are calculated separately using Pearson's similarity metric function：
Sim (target user, user 1)=- 0.76
Sim (target user, user 3)=- 0.53
If choosing adjacent user of the candidate adjacent user of highest two of similarity as target user, the neighbour of target user
Nearly user NTarget user(2)={ user 3, and user 1 }；
Step 5. article that do not score target user carries out prediction scoring：
According to the adjacent user of target userN au (k)Score information, using adjustment weighted sum function (6) to target userau
The article set not scoredPrediction scoring is carried out, the prediction grade form of target user is obtained；
(6)
In formula (6),P au,i Indicate target userauTo articleiPrediction scoring,U i ={u∈U|r u,i ≠*}Object was evaluated in expression
ProductiUser set；λAs a regularization factors：
(7)
According to the score information of the adjacent user of target user, the article 5 not scored target user using adjustment weighting function
Prediction scoring is carried out with article 6, it is as a result as follows：
PTarget user, article 5= 2.16
PTarget user, article 6= 4.93
Step 6. chooses the prediction highest article of score value as recommendation results, and article 6 will recommend target user.
Experiment：
(1) public data collection is used
The public data collection MovieLens of test commending system performance is often used in using commending system field.The data set packet
Containing 943 users, 1682 films and 100000 scorings, score value are distributed as { 1,2,3,4,5 }, and each user is at least to 20
A film is scored.
(2) evaluation measurement
The present invention is come the accuracy of metric algorithm, MAE and RMSE using mean absolute error MAE and root-mean-square error RMSE
Measure the accuracy of recommendation results by calculating the deviation between the practical scoring of user and prediction scoring, therefore, MAE and
RMSE is smaller, recommends precision higher：
(8)
(9)
Using coverage Coverage come the diversity of metric algorithm.Coverage refers to recommend the type of goods of target user
It is the ratio for evaluating type of goods to account for all target users, and therefore, coverage Coverage is higher, is recommended more diversified.
(10)
Formula（10）In,, wherein Su,iThe adjacent user of the user u of article i was evaluated in expression.
（3）Parameter setting
The present invention using Pearson's similarity metric function calculate user similarity, using adjustment weighting function to target user not
The article of scoring carries out prediction scoring.The favorite the first two goods attribute of target user is chosen to like belonging to as target user
Property.In order to clearly compare the present invention with legacy user's collaborative filtering, target user adjacent user quantity K ∈ 20,
25,30,…,60}.To recommend the quantity set of article is { 2,4,6,8,10,12 }.
（4）Experimental result compares and analysis
The present invention the personalized user collaborative filtering recommending method based on Covering reduct indicated with CBCF, traditional user collaborative
Filter algorithm indicates that Fig. 2 shows the result of precision metrology MAE and RMSE with UBCF.By Fig. 2 data it is found that with mesh
The increase of adjacent user's number of user is marked, MAE the and RMSE results of CBCF algorithms are always less than the result of UBCF algorithms.Due to
MAE and RMSE are smaller, recommend precision higher, therefore CBCF can recommend article more higher than UBCF precision.Fig. 3 shows various
Property measurement Coverage's as a result, as shown in Figure 3, with the increase of adjacent user's number of target user, CBCF algorithms
Coverage is significantly greater than the coverage of UBCF.Since coverage Coverage is higher, recommendation is more diversified, therefore CBCF can be pushed away
Recommend the article more various than UBCF.Comprehensive Experiment result is it is found that the present invention can provide high-precision and diversified push away simultaneously
It recommends as a result, to realize the personalized recommendation of target user.
Claims (1)
1. a kind of personalized user collaborative filtering recommending method based on Covering reduct, it is characterised in that successively in accordance with the following steps
It carries out：
Step 1. statistics forms two-dimentional score information table：
Two-dimentional score information table is formed to the score information of article according to userRM={ U, I, R ∪ { * } }；The two dimension scoring letter
Cease tableRMIn,UIndicate the set of user,IIndicate the set of article,R∪{*}Indicate scoring set of the user to article, wherein*
Indicate that user does not score to article；
Enable useru∈UTo articlei∈IScoring ber u,i ∈R∪{*}, and useruAverage score be；θIt is commented for user
The threshold value divided, ifr u,i ≥θ, show useruLike articlei；UseruThe article collection to have scored is combined intoI u ={i∈I|r u,i ≠*}；For useruThe article set not scored；Goods attribute matrix isAM；In user's set U, liked if there is user a
The article set of love is contained in the favorite article set of user b, then user a is known as the redundant subscribers of target user；
Step 2. carries out yojan using Covering reduct algorithm to redundant subscribers：
Step 2.1 enables article setIAs domainI, in domainIIn, the article that each user likes forms a set；In object
Product attribute matrixAMMiddle extraction target user's likes attribute：
(1)
In formula (1),mIndicate the number of attribute,at m Indicate an attribute,av m Indicate attributeat m Value；
Step 2.2. likes attribute using the target user of acquisition, builds the decision set of target userD, decision setDBy having
The article collection of attribute is liked to be combined into：
D={ i ∈ I|at 1(i)= av 1 ,at 2 (i)=av 2 ,…,at m (i)=av m (2)
In formula (2),at m (i)= av m Indicate articleiIn attributeat m On value beav m ；
Step 2.3. is by domainIThe decision set of target user is reduced to by article setD, i.e. domainD；For each useru ∈U, build useruIn domainDOn like article setC u ：
(3)
It enablesC*=D-∪C u , C={ C 1 ,C 2 …C n , C*} Target user is constituted in domainDOne coveringC；
Step 2.4. utilizes Covering reduct algorithm, by redundant elements from coveringCMiddle yojan obtains the covering after yojanreduct(C) and yojan after userU r :
(4)
Step 3. utilizes the user after yojanU r Build target userauCandidate adjacent useru；
Step 4. calculates the similarity of target user and candidate adjacent user, chooses the adjacent user of target user：
Target user is calculated using Pearson's similarity metric function (5)auWith candidate adjacent useru∈U r Between similarity,
(5)
In formula (5),sim(au,u)Indicate target userauWith candidate adjacent useru∈U r Between similarity,I au ={i∈I| r au,i ≠*}Indicate target userauThe article set evaluated,Indicate the average score value of target user；
Then before selection similarity is highKAdjacent user of a candidate adjacent user as target userN au (k)；
Step 5. article that do not score target user carries out prediction scoring：
According to the adjacent user of target userN au (k)Score information, using adjustment weighted sum function (6) to target userauNot
The article set of scoringPrediction scoring is carried out, the prediction grade form of target user is obtained；
(6)
In formula (6),P au,i Indicate target userauTo articleiPrediction scoring,U i ={u∈U|r u,i ≠*}Article was evaluated in expressioniUser set；λAs a regularization factors：
(7)
Before step 6 selection prediction scorings are highNA article is as recommendation results.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810486715.0A CN108549729B (en) | 2018-05-21 | 2018-05-21 | Personalized user collaborative filtering recommendation method based on coverage reduction |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810486715.0A CN108549729B (en) | 2018-05-21 | 2018-05-21 | Personalized user collaborative filtering recommendation method based on coverage reduction |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108549729A true CN108549729A (en) | 2018-09-18 |
| CN108549729B CN108549729B (en) | 2021-06-11 |
Family
ID=63495385
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810486715.0A Active CN108549729B (en) | 2018-05-21 | 2018-05-21 | Personalized user collaborative filtering recommendation method based on coverage reduction |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108549729B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110990689A (en) * | 2019-11-11 | 2020-04-10 | 辽宁师范大学 | Collaborative filtering recommendation algorithm based on rough set rule extraction |
| CN112612967A (en) * | 2020-09-11 | 2021-04-06 | 辽宁师范大学 | Collaborative filtering recommendation method based on link clustering and reduction |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090006398A1 (en) * | 2007-06-29 | 2009-01-01 | Shing Yan Lam | Recommendation system with multiple integrated recommenders |
| CN102662975A (en) * | 2012-03-12 | 2012-09-12 | 浙江大学 | Bidirectional and clustering mixed friend recommendation method |
| CN102982107A (en) * | 2012-11-08 | 2013-03-20 | 北京航空航天大学 | Recommendation system optimization method with information of user and item and context attribute integrated |
| CN106934071A (en) * | 2017-04-27 | 2017-07-07 | 北京大学 | Recommendation method and device based on Heterogeneous Information network and Bayes's personalized ordering |
-
2018
- 2018-05-21 CN CN201810486715.0A patent/CN108549729B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090006398A1 (en) * | 2007-06-29 | 2009-01-01 | Shing Yan Lam | Recommendation system with multiple integrated recommenders |
| CN102662975A (en) * | 2012-03-12 | 2012-09-12 | 浙江大学 | Bidirectional and clustering mixed friend recommendation method |
| CN102982107A (en) * | 2012-11-08 | 2013-03-20 | 北京航空航天大学 | Recommendation system optimization method with information of user and item and context attribute integrated |
| CN106934071A (en) * | 2017-04-27 | 2017-07-07 | 北京大学 | Recommendation method and device based on Heterogeneous Information network and Bayes's personalized ordering |
Non-Patent Citations (1)
| Title |
|---|
| ZHIPENG ZHANG ET AL.: "Neighbor selection for user-based collaborative filtering using covering-based rough sets", 《ANNALS OF OPERATIONS RESEARCH》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110990689A (en) * | 2019-11-11 | 2020-04-10 | 辽宁师范大学 | Collaborative filtering recommendation algorithm based on rough set rule extraction |
| CN110990689B (en) * | 2019-11-11 | 2023-04-18 | 辽宁师范大学 | Collaborative filtering recommendation algorithm based on rough set rule extraction |
| CN112612967A (en) * | 2020-09-11 | 2021-04-06 | 辽宁师范大学 | Collaborative filtering recommendation method based on link clustering and reduction |
| CN112612967B (en) * | 2020-09-11 | 2023-07-18 | 辽宁师范大学 | Collaborative filtering recommendation method based on link clustering and reduction |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108549729B (en) | 2021-06-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104281956B (en) | The dynamic recommendation method for adapting to user interest variation based on temporal information | |
| CN103514304B (en) | Project recommendation method and device | |
| CN104966125B (en) | A kind of article scoring of social networks and recommend method | |
| CN107122390B (en) | Recommendation system construction method based on user group | |
| CN107633444B (en) | Recommendation system noise filtering method based on information entropy and fuzzy C-means clustering | |
| KR101620748B1 (en) | Item recommendation method and apparatus | |
| JP6261547B2 (en) | Determination device, determination method, and determination program | |
| CN108804683A (en) | Associate(d) matrix decomposes and the film of collaborative filtering recommends method | |
| CN107894998B (en) | Video recommendation method and device | |
| CN103559622A (en) | Characteristic-based collaborative filtering recommendation method | |
| CN106471491A (en) | A kind of collaborative filtering recommending method of time-varying | |
| CN107180093A (en) | Information search method and device and ageing inquiry word recognition method and device | |
| CN108132964B (en) | Collaborative filtering method for scoring project classes based on user | |
| CN108334592A (en) | A kind of personalized recommendation method being combined with collaborative filtering based on content | |
| CN104657336B (en) | A kind of personalized recommendation method based on half cosine function | |
| CN109977299B (en) | Recommendation algorithm fusing project popularity and expert coefficient | |
| CN106776859A (en) | Mobile solution App commending systems based on user preference | |
| CN109902235A (en) | User preference based on bat optimization clusters Collaborative Filtering Recommendation Algorithm | |
| Yu et al. | A coupled clustering approach for items recommendation | |
| CN104008193B (en) | A kind of information recommendation method based on group of typical user discovery technique | |
| CN105809275A (en) | Item scoring prediction method and apparatus | |
| CN109271592A (en) | A kind of item recommendation method, device, storage medium and program product | |
| CN108549729A (en) | Personalized user collaborative filtering recommending method based on Covering reduct | |
| CN109508407A (en) | The tv product recommended method of time of fusion and Interest Similarity | |
| CN106021428A (en) | KNN and three-way decision-based movie recommendation method |
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 |