CN110297939A - A kind of music personalization system of fusion user behavior and cultural metadata - Google Patents
A kind of music personalization system of fusion user behavior and cultural metadata Download PDFInfo
- Publication number
- CN110297939A CN110297939A CN201910540276.1A CN201910540276A CN110297939A CN 110297939 A CN110297939 A CN 110297939A CN 201910540276 A CN201910540276 A CN 201910540276A CN 110297939 A CN110297939 A CN 110297939A
- Authority
- CN
- China
- Prior art keywords
- user
- song
- music
- singer
- songs
- 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
Links
- 230000004927 fusion Effects 0.000 title claims abstract description 11
- 235000019640 taste Nutrition 0.000 claims abstract description 8
- 238000012937 correction Methods 0.000 claims abstract description 7
- 230000006399 behavior Effects 0.000 claims description 23
- 239000011159 matrix material Substances 0.000 claims description 15
- 238000011156 evaluation Methods 0.000 claims description 12
- 238000004364 calculation method Methods 0.000 claims description 9
- HUTDUHSNJYTCAR-UHFFFAOYSA-N ancymidol Chemical compound C1=CC(OC)=CC=C1C(O)(C=1C=NC=NC=1)C1CC1 HUTDUHSNJYTCAR-UHFFFAOYSA-N 0.000 claims description 6
- 230000000694 effects Effects 0.000 claims description 4
- 238000000034 method Methods 0.000 claims description 4
- 238000001914 filtration Methods 0.000 claims description 3
- 239000011521 glass Substances 0.000 claims description 3
- 230000011273 social behavior Effects 0.000 claims description 3
- 230000009286 beneficial effect Effects 0.000 abstract description 2
- 230000001737 promoting effect Effects 0.000 abstract description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 241000270322 Lepidosauria Species 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000003930 cognitive ability Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000008451 emotion Effects 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 210000004209 hair Anatomy 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/60—Information retrieval; Database structures therefor; File system structures therefor of audio data
- G06F16/63—Querying
- G06F16/635—Filtering based on additional data, e.g. user or group profiles
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/60—Information retrieval; Database structures therefor; File system structures therefor of audio data
- G06F16/65—Clustering; Classification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/60—Information retrieval; Database structures therefor; File system structures therefor of audio data
- G06F16/68—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
- G06F16/686—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using information manually generated, e.g. tags, keywords, comments, title or artist information, time, location or usage information, user ratings
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Multimedia (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Library & Information Science (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The purpose of the present invention is to provide the music personalization systems of a kind of fusion user behavior and cultural metadata, including user behavior cluster module, cultural metadata cluster module, music personalized recommendation module and music to recommend correction module.The beneficial effects of the invention are as follows the musical tastes that present system can respect fully user individual, the cultural metadata feature of music can be merged from user perspective, from diversity, several angles such as accuracy and pleasantly surprised degree help user to find more high-quality music elements in music cloud platform vast resources, user can stroll, sound wherein, the utilization rate of music sources in music cloud platform is constantly promoted while promoting user satisfaction.
Description
Technical field
The invention belongs to music technology fields, are related to the music personalization system of a kind of fusion user behavior and cultural metadata
System.
Background technique
With the continuous development application of information technology and mobile terminal, problem of information overload more and more attention has been paid to.List is just
For song, musical works library forms the music information library of magnanimity.For music-lover, how from the music information of magnanimity
Oneself interested music is easily and efficiently found in library becomes more and more difficult;Simultaneously for the creator of music, how to allow
The music of oneself creation can show one's talent from the music libraries of magnanimity, quickly be obtained by people and one and difficult
Thing.For this problem, music personalized recommendation system comes into being, and individualized music recommender system to people by providing
With targetedly musical works, so as to alleviate problem of information overload.Fan of passively holding music from the beginning searches
For Suo Bianwei actively according to the demand of music-lover, individualized music recommender system obtains the attention of people.It realizes accurately
Personalized recommendation system realizes accurate music push it is necessary to comprehensively consider the correlation between the demand of user and song.Music
As the affective medium of people, per se with the emotion of the extremely strong mankind, so the metadata of music will be comprehensively considered,
The information such as comment, the label of style, song including song, in this way when carrying out music push, so that it may according to these
The behavior of information combination user precisely pushes.There is cold start-up, Sparse in existing music recommender system, music is recommended
Accuracy rate it is low, lacking individuality.The present invention comprehensively consider user behavior and user locating for environment, respect fully user's
Hobby and individual demand can timely and effectively recommend the individualized music for being suitble to its local environment feature for user, have
Context aware function.
Summary of the invention
The purpose of the present invention is to provide the music personalization system of a kind of fusion user behavior and cultural metadata, this hairs
Bright beneficial effect is the musical taste that present system can respect fully user individual, can be come from user perspective
Merge the cultural metadata feature of music.User is helped to find music from several angles such as diversity, accuracy and pleasantly surprised degree
More high-quality music elements in cloud platform vast resources, user can stroll, sound wherein, full promoting user
The utilization rate of music sources in music cloud platform is constantly promoted while meaning is spent.
The technical scheme adopted by the invention is that including user behavior cluster module, cultural metadata cluster module, music
Personalized recommendation module and music recommend correction module.
Further, user behavior cluster module includes
(1), singer's preference degree SPD of user:
SPD is used to describe user to the level of interest of singer, is some singer if user misses potter some singer
Song pursuer, then be just likely to all can be interested be in all songs of this singer by user, it is possible to form one
A individual singer's list, the division for special user to song calculate different users to the hobby of different songs
It spends SPD (p, q), calculation formula is as follows:
Wherein SPD (p, q) indicates some user p to the preference degree of singer q, and sumSongCount (p) indicates user's concern
All list of songs, that is, all song information data of the user acquired, songCount (p, q) indicate user concern institute
There is the number of songs containing singer q in song, can thus define user to the preference degree SPD of specific singer;
(2), preference degree SLP of the user to music language:
Wherein SLP (p, q) represents user p to the musical taste degree of song language q, and songLan (p, q) indicates user's song
In list, belong to the number of songs of language q;
(3), user is to music label preference degree SLD:
Wherein, SLD (p, q) is musical taste degree of the user p to song features q, and songLab (p, q) is to indicate p pairs of user
Belong to the total number of labels of q in song label, sumLabCount (p) is total number of labels of the user to all songs.
Further, it is that many music based on label are personalized that cultural metadata cluster module, which is the cultural metadata of music,
The data basis of recommender system analyzes the relevance between singer using the label that user provides, by acquiring user
The song comment information delivered on music platform clusters song using deep learning clustering algorithm, by similar song
It clusters under a module, makes full use of the Social behaviors of user.Tracking activity of the user on social networks can be preferably
Obtain the demand of user, thus it is speculated that the preference of user, single user describe file by considering that user and Web Community good friend's is mutual
It is dynamic to be expanded.
Further, music personalized recommendation module utilizes improved music personalized recommendation collaborative filtering, carries out a
Propertyization is recommended, and is constructed a rating matrix to the downloading of song, collection, broadcasting time, splitting glass opaque according to user, is passed through this
Matrix calculates the arest neighbors of user, realizes the song recommendations to user:
(1) song rating matrix is constructed
It is scored song by user the operation of song, by the numerical value of these indexs, is filled into what needs constructed
Among rating matrix, the arest neighbors between user can be calculated by this matrix;
(2) similarity between user is calculated
Using improved cosine similarity amount method, the similarity between the user to be recommended and other users is calculated,
Define user U1And U2For the operation I of same songijIt indicates, IiAnd IjTable is divided to indicate user U1And U2To respective song
Operation, user U is calculated by following formula1And U2Between similarity:
(3) song recommendations result is generated
By calculating similarity, the song of similar users in the neighbour space of the user of quasi- recommendation is obtained, applied forecasting is commented
The formula divided, calculates scoring of the quasi- recommended user for song, then the selection scoring higher song of score ratio from list of songs
Song recommends user, prediction scoring calculation formula:
Wherein, most like neighbour's collection of user sharesNBSuTo indicate.
Further, music recommends correction module by the appraisement system of personalized recommendation to the personalized recommendation formed
User is modified, and continuously improves, and in evaluation index, the accuracy of recommendation is used to measure personalized recommendation system or recommendation
Algorithm predicts the ability of user behavior, is evaluation index the most universal and the most basic, using absolute error MAE and mean square error
Poor RMSE is estimated, and calculation method is as follows:
Wherein, Ri,jThe user i of actual demand for to(for) song j,It is that user i recommends the demand of j pre- for recommender system
It surveys, N is the value of predicted quantity.
Detailed description of the invention
Fig. 1 is the music personalized recommendation system overall framework figure of fusion user behavior and cultural metadata of the invention.
Specific embodiment
The present invention is described in detail With reference to embodiment.
Fig. 1 is the music personalization system frame diagram for merging user behavior and cultural metadata.Present system includes using
Family behavior cluster module, cultural metadata cluster module, music personalized recommendation module and music recommend correction module.
1, user behavior cluster module: user is the emphasis for realizing personalized recommendation, only by the hobby of user and demand
Holding preferably to realize the push of precision music to user.The present invention utilizes reptile instrument, crawls from music cloud platform
The song list of user oneself setting and collection, and user gets off to the operation note of song, user preference vector model is established, it is fixed
The following concept of justice.
(1), singer's preference degree SPD (Singer Preference Degree) of user
SPD is used to describe user to the level of interest of singer.It is some singer if user misses potter some singer
" song pursuer ", then be just likely to all can be interested be in all songs of this singer by user.It is possible to be formed
One individual singer's list, the division for special user to song calculate happiness of the different users to different songs
Spend SPD (p, q) well.Calculation formula is as follows:
Wherein SPD (p, q) indicates some user p to the preference degree of singer q, and sumSongCount (p) indicates user's concern
All list of songs, that is, all song information data of the user acquired.SongCount (p, q) indicates the institute of user's concern
There is the number of songs containing singer q in song, can thus define user to the preference degree SPD of specific singer.
(2), preference degree SLP (Song Language Preference) of the user to music language
Passing through the single analysis of the music song to a large number of users, it has been found that the user having compares hobby for English song,
It is special to establish a song list for English song, while the user yet having does not like English song, and like Chinese songs.According to
This case, we calculate the happiness of user according to the classification of Chinese, English, four kinds of America and Europe, Japan and Korea S songs in music cloud platform
Spend SLP well.
Wherein SLP (p, q) represents user p to the musical taste degree of song language q.SongLan (p, q) indicates user's song
In list, belong to the number of songs of language q, user can be obtained according to this formula to the preference degree of music language.
(3), user is to music label preference degree SLD (Singer Label Degree)
For fixed song, we cannot be changed the information of song, but we can pass through user
The relevant information of this song song is calculated to the preference of song.For song, the internal characteristics of itself are to change
It does not become, is determined by the concern behavior of user.Different users also has different labels for same first song.Cause
This, can obtain user to the preference degree of different songs according to user to the markup information of song.
Therefore it can establish, based on user to the preference degree SLD of song label.
Wherein, SLD (p, q) is musical taste degree of the user p to song features q, and songLab (p, q) is to indicate p pairs of user
Belong to the total number of labels of q in song label, sumLabCount (p) is total number of labels of the user to all songs.
An interest table can be established according to user to the fancy grade of song, indicate to use with the affective characteristics of song
The affective characteristics interest table at family, table 1 are user feeling feature interest table.
Table 1
2, cultural metadata cluster module:
The cultural metadata of music is the data basis of many music personalized recommendation systems based on label, utilizes user
The label of offer analyzes the relevance between singer.Present people inherently carry social attribute, are used by acquisition
The song comment information that family is delivered on music platform clusters song, will be under similar song clusters a to module.
Make full use of the Social behaviors of user.The demand of user can preferably be obtained by tracking activity of the user on social networks, be pushed away
Survey the preference of user.Single user describes file (User Profiles) can be by considering user and Web Community good friend's
Interaction is expanded, such as the music between good friend is recommended, shares or collected.
3, music personalized recommendation module:
Using improved music personalized recommendation collaborative filtering, personalized recommendation is carried out.It is sung in antiphonal style first according to user
The downloading of song, collection, broadcasting time, splitting glass opaque construct a rating matrix, can calculate user's by this matrix
Arest neighbors realizes the song recommendations to user.
(1) song rating matrix is constructed
Operation by user to song, such as download time, broadcasting time, four uncles of collection and the comment number of song etc.
The operation of user, scores to song, by the numerical value of these indexs, is filled among the rating matrix for needing to construct, passes through
This matrix can calculate the arest neighbors between user.
(2) similarity between user is calculated
The present invention uses improved cosine similarity amount method, calculates between the user and other users that we to be recommended
Similarity.Define user U1And U2For the operation I of same songijIt indicates, IiAnd IjTable is divided to indicate user U1And U2It is right
The operation of respective song, then user U can be calculated by following formula1And U2Between similarity.
(3) song recommendations result is generated
By calculating similarity, the song of similar users in the neighbour space of the user of our available quasi- recommendations can
With the formula that applied forecasting is scored, scoring of the quasi- recommended user for song is calculated, then the selection scoring point from list of songs
The relatively high song recommendations of number are to user.Prediction scoring calculation formula:
Wherein, most like neighbour's collection of user sharesNBSuTo indicate.
4, music recommends correction module:
The personalized recommendation user formed is modified by the appraisement system of personalized recommendation, is constantly changed
Into.
The effect that evaluation index evaluation is recommended, evaluation index is a kind of a kind of mode for evaluating recommender system, final to use
Family is only the user of recommender system, and the quality that a recommender system is only evaluated from the angle of evaluation index is that do not have
Meaning, but we can be by evaluation index come the quality of one recommender system of Primary Evaluation, in order to form higher push away
The system of recommending is made efforts.
In evaluation index, the accuracy of recommendation is used to measure personalized recommendation system or proposed algorithm prediction user behavior
Ability, be the most universal and the most basic evaluation index.Estimated using absolute error MAE and mean square error RMSE,
Calculation method is as follows:
Wherein, Ri,jThe user i of actual demand for to(for) song j,It is that user i recommends the demand of j pre- for recommender system
It surveys, N is the value of predicted quantity.Compared with MAE, RMSE is punished using square root, can obtain more rigorous and accurately result.
MAE and RMSE is lower to illustrate that system prediction precision is higher.In actual use, because user has the cognitive ability of project
Limit, so N is often limited to some specific integer.
The above is only not to make limit in any form to the present invention to better embodiment of the invention
System, any simple modification that embodiment of above is made according to the technical essence of the invention, equivalent variations and modification,
Belong in the range of technical solution of the present invention.
Claims (5)
1. the music personalization system of a kind of fusion user behavior and cultural metadata, it is characterised in that: poly- including user behavior
Generic module, cultural metadata cluster module, music personalized recommendation module and music recommend correction module.
2. according to the music personalization system of a kind of fusion user behavior and cultural metadata described in claim 1, feature exists
In: the user behavior cluster module includes
(1), singer's preference degree SPD of user:
SPD is used to describe user to the level of interest of singer, is the song of some singer if user misses potter some singer
Bent pursuer, then be just likely to all can be interested be in all songs of this singer by user, it is possible to form a list
Only singer's list, the division for special user to song calculate different users to the preference degree SPD of different songs
(p, q), calculation formula are as follows:
Wherein SPD (p, q) indicates some user p to the preference degree of singer q, and sumSongCount (p) indicates the institute of user's concern
There is list of songs, that is, all song information data of the user acquired, songCount (p, q) indicates all songs of user's concern
Number of songs containing singer q in song can thus define user to the preference degree SPD of specific singer;
(2), preference degree SLP of the user to music language:
Wherein SLP (p, q) represents user p to the musical taste degree of song language q, and songLan (p, q) indicates user's list of songs
In, belong to the number of songs of language q;
(3), user is to music label preference degree SLD:
Wherein, SLD (p, q) is musical taste degree of the user p to song features q, and songLab (p, q) is to indicate user p to song
Belong to the total number of labels of q in label, sumLabCount (p) is total number of labels of the user to all songs.
3. according to the music personalization system of a kind of fusion user behavior and cultural metadata described in claim 1, feature exists
In: the culture metadata cluster module is that the cultural metadata of music is many music personalized recommendation systems based on label
Data basis, the relevance between singer is analyzed using the label that user provides, it is flat in music by acquisition user
The song comment information delivered on platform clusters song using Single-Pass algorithm, by similar song clusters to one
Under a module, the Social behaviors of user are made full use of.User can preferably be obtained by tracking activity of the user on social networks
Demand, thus it is speculated that the preference of user, single user describe file by considering that user is expanded with interacting for Web Community good friend
Exhibition.
4. according to the music personalization system of a kind of fusion user behavior and cultural metadata described in claim 1, feature exists
In: the music personalized recommendation module utilizes improved music personalized recommendation collaborative filtering, carries out personalized recommendation,
One rating matrix is constructed to the downloading of song, collection, broadcasting time, splitting glass opaque according to user, is calculated by this matrix
The arest neighbors of user out realizes the song recommendations to user:
(1) song rating matrix is constructed
It is scored song by user the operation of song, by the numerical value of these indexs, is filled into the scoring for needing to construct
Among matrix, the arest neighbors between user can be calculated by this matrix;
(2) similarity between user is calculated
Using improved cosine similarity amount method, the similarity between the user to be recommended and other users is calculated, is defined
User U1And U2For the operation I of same songijIt indicates, IiAnd IjTable is divided to indicate user U1And U2To the behaviour of respective song
Make, user U is calculated by following formula1And U2Between similarity:
(3) song recommendations result is generated
By calculating similarity, the song of similar users in the neighbour space of the user of quasi- recommendation is obtained, applied forecasting scoring
Formula calculates scoring of the quasi- recommended user for song, then the selection scoring higher song of score ratio pushes away from list of songs
It recommends to user, prediction scoring calculation formula:
Wherein, most like neighbour's collection of user sharesNBSuTo indicate.
5. according to the music personalization system of a kind of fusion user behavior and cultural metadata described in claim 1, feature exists
In: the music recommends correction module to carry out by the appraisement system of personalized recommendation to the personalized recommendation user formed
Amendment, continuously improves, and in evaluation index, the accuracy of recommendation is used to measure personalized recommendation system or proposed algorithm prediction
The ability of user behavior is the most universal and the most basic evaluation index, using absolute error MAE and mean square error RMSE into
Row is estimated, and calculation method is as follows:
Wherein, Ri,jThe user i of actual demand for to(for) song j,It is the requirement forecasting that user i recommends j, N for recommender system
It is the value of predicted quantity.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910540276.1A CN110297939A (en) | 2019-06-21 | 2019-06-21 | A kind of music personalization system of fusion user behavior and cultural metadata |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910540276.1A CN110297939A (en) | 2019-06-21 | 2019-06-21 | A kind of music personalization system of fusion user behavior and cultural metadata |
Publications (1)
Publication Number | Publication Date |
---|---|
CN110297939A true CN110297939A (en) | 2019-10-01 |
Family
ID=68028408
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910540276.1A Pending CN110297939A (en) | 2019-06-21 | 2019-06-21 | A kind of music personalization system of fusion user behavior and cultural metadata |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110297939A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111507804A (en) * | 2020-04-21 | 2020-08-07 | 莫毓昌 | Emotion perception commodity recommendation method based on mixed information fusion |
CN113094542A (en) * | 2021-03-24 | 2021-07-09 | 西安交通大学 | Set ordering music recommendation method aiming at user implicit feedback data |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102402625A (en) * | 2011-12-28 | 2012-04-04 | 深圳市五巨科技有限公司 | Method and system for recommending music |
CN104008138A (en) * | 2014-05-08 | 2014-08-27 | 南京邮电大学 | Music recommendation method based on social network |
CN104731954A (en) * | 2015-04-01 | 2015-06-24 | 天翼爱音乐文化科技有限公司 | Music recommendation method and system based on group perspective |
CN104899302A (en) * | 2015-06-10 | 2015-09-09 | 百度在线网络技术(北京)有限公司 | Method and device for recommending music to user |
CN106599114A (en) * | 2016-11-30 | 2017-04-26 | 上海斐讯数据通信技术有限公司 | Music recommendation method and system |
CN108304441A (en) * | 2017-11-14 | 2018-07-20 | 腾讯科技(深圳)有限公司 | Network resource recommended method, device, electronic equipment, server and storage medium |
CN109063163A (en) * | 2018-08-14 | 2018-12-21 | 腾讯科技(深圳)有限公司 | A kind of method, apparatus, terminal device and medium that music is recommended |
-
2019
- 2019-06-21 CN CN201910540276.1A patent/CN110297939A/en active Pending
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102402625A (en) * | 2011-12-28 | 2012-04-04 | 深圳市五巨科技有限公司 | Method and system for recommending music |
CN104008138A (en) * | 2014-05-08 | 2014-08-27 | 南京邮电大学 | Music recommendation method based on social network |
CN104731954A (en) * | 2015-04-01 | 2015-06-24 | 天翼爱音乐文化科技有限公司 | Music recommendation method and system based on group perspective |
CN104899302A (en) * | 2015-06-10 | 2015-09-09 | 百度在线网络技术(北京)有限公司 | Method and device for recommending music to user |
CN106599114A (en) * | 2016-11-30 | 2017-04-26 | 上海斐讯数据通信技术有限公司 | Music recommendation method and system |
CN108304441A (en) * | 2017-11-14 | 2018-07-20 | 腾讯科技(深圳)有限公司 | Network resource recommended method, device, electronic equipment, server and storage medium |
CN109063163A (en) * | 2018-08-14 | 2018-12-21 | 腾讯科技(深圳)有限公司 | A kind of method, apparatus, terminal device and medium that music is recommended |
Non-Patent Citations (1)
Title |
---|
周永章,等: "《地球科学大数据挖掘与机器学习》", 广州:中山大学出版社, pages: 110 - 111 * |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111507804A (en) * | 2020-04-21 | 2020-08-07 | 莫毓昌 | Emotion perception commodity recommendation method based on mixed information fusion |
CN111507804B (en) * | 2020-04-21 | 2022-05-13 | 莫毓昌 | Emotion perception commodity recommendation method based on mixed information fusion |
CN113094542A (en) * | 2021-03-24 | 2021-07-09 | 西安交通大学 | Set ordering music recommendation method aiming at user implicit feedback data |
CN113094542B (en) * | 2021-03-24 | 2023-08-15 | 西安交通大学 | Set ordering music recommendation method for implicit feedback data of user |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110574387B (en) | Recommending live streaming content using machine learning | |
Christensen et al. | Entertainment recommender systems for group of users | |
Di Noia et al. | Adaptive multi-attribute diversity for recommender systems | |
Casadei et al. | Global cities, creative industries and their representation on social media: A micro-data analysis of Twitter data on the fashion industry | |
Pavlidis | Recommender systems, cultural heritage applications, and the way forward | |
Boratto et al. | Discovery and representation of the preferences of automatically detected groups: Exploiting the link between group modeling and clustering | |
CN110402438A (en) | Music from focus inquiry is recommended | |
CN107301247B (en) | Method and device for establishing click rate estimation model, terminal and storage medium | |
Schedl et al. | Putting the User in the Center of Music Information Retrieval. | |
WO2009126815A2 (en) | Diversified, self-organizing map system and method | |
Beer et al. | The hidden dimensions of the musical field and the potential of the new social data | |
CN110069713A (en) | A kind of personalized recommendation method based on user's context perception | |
Kim et al. | Exploring characteristics of video consuming behaviour in different social media using K-pop videos | |
Keller et al. | Recommender systems for museums: evaluation on a real dataset | |
CN110297939A (en) | A kind of music personalization system of fusion user behavior and cultural metadata | |
Said | Evaluating the accuracy and utility of recommender systems | |
Dixit et al. | Weighted percentile-based context-aware recommender system | |
Valera et al. | Context-aware music recommender systems for groups: A comparative study | |
Elahi | Empirical evaluation of active learning strategies in collaborative filtering | |
Zhang et al. | Learning to build accurate service representations and visualization | |
Vasu et al. | Music Information Retrieval Using Similarity Based Relevance Ranking Techniques | |
Zhang et al. | Musical preference in an online music community in China | |
Wu et al. | Result Diversification in Search and Recommendation: A Survey | |
KR20170123660A (en) | Algorithm radio for arbitrary text queries | |
KR101098870B1 (en) | Posts Search Method and Apparatus Based on User Rank Similarity |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20191001 |
|
RJ01 | Rejection of invention patent application after publication |