CN104008138B - A kind of music based on social networks recommends method - Google Patents
A kind of music based on social networks recommends method Download PDFInfo
- Publication number
- CN104008138B CN104008138B CN201410192981.4A CN201410192981A CN104008138B CN 104008138 B CN104008138 B CN 104008138B CN 201410192981 A CN201410192981 A CN 201410192981A CN 104008138 B CN104008138 B CN 104008138B
- Authority
- CN
- China
- Prior art keywords
- user
- song
- music
- friend
- recommended
- 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.)
- Expired - Fee Related
Links
- 238000000034 method Methods 0.000 title claims description 9
- 238000007621 cluster analysis Methods 0.000 claims description 6
- 238000010224 classification analysis Methods 0.000 claims description 5
- 238000011156 evaluation Methods 0.000 claims description 4
- 230000005012 migration Effects 0.000 claims description 4
- 238000013508 migration Methods 0.000 claims description 4
- 230000009471 action Effects 0.000 claims description 2
- 238000007418 data mining Methods 0.000 abstract description 9
- 238000005516 engineering process Methods 0.000 abstract description 8
- 238000013461 design Methods 0.000 abstract description 5
- 201000004569 Blindness Diseases 0.000 abstract description 4
- 230000008859 change Effects 0.000 abstract description 4
- 230000011218 segmentation Effects 0.000 abstract description 3
- 230000001360 synchronised effect Effects 0.000 abstract 1
- 230000008569 process Effects 0.000 description 2
- 241000238557 Decapoda Species 0.000 description 1
- 235000010627 Phaseolus vulgaris Nutrition 0.000 description 1
- 244000046052 Phaseolus vulgaris Species 0.000 description 1
- 238000009412 basement excavation Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 239000011435 rock Substances 0.000 description 1
- 230000009897 systematic effect Effects 0.000 description 1
- 230000007704 transition 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
- G06F16/90—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
- 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)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Library & Information Science (AREA)
- Multimedia (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The present invention is a kind of individualized music suggested design.For blindness and mechanical problem that traditional music is recommended, using data mining technology, have levels by subscriber segmentation, according to the historical record of user, analyze a variety of demands.It is combined with social networks simultaneously, the music and disagreeable music liked according to user are clustered to user, and to each user generation interest tendency label, recommended user adds as a friend neighboring user, and with the change synchronous change good friend of user interest.Then with friend circle as data set, recommended using correlation rule so that recommend more accurate efficient.Cloud computing technology is used simultaneously, to solve the music information and user data of magnanimity, with good expansion.
Description
Technical field
The present invention is a kind of individualized music suggested design.The blindness recommended for traditional music and mechanical ask
Topic, using data mining technology, have levels by subscriber segmentation, and be combined with social networks so that recommend more accurate high
Effect.Cloud computing technology is used simultaneously, to solve the music information and user data of magnanimity.Belong to data mining and cloud computing neck
Domain.
Background technology
With developing rapidly for Internet technologies, music recommended website miscellaneous is arisen at the historic moment, network digital sound
Pleasure is increasingly becoming indispensable important component in people's studying and living.Brand-new music carrier is being music site
While business opportunity is provided, it was also proposed that new challenge.How to attract new user in network world, old user kept here, as sound
The main task of happy website.On the other hand, user will therefrom find oneself requirement in the digital music in face of magnanimity
Just like looking for a needle in a haystack.It would therefore be desirable to the music tendentiousness of the comprehensive understanding user of energy, reflects user interest comprehensively,
And the music commending system of user's potential demand can be excavated.
Music class product have passed through three development epoch, and first epoch is the program request epoch, user by song title and
Singer comes the song that program request oneself is wanted to listen, such as dried shrimp；Second epoch is that data mining is carried out by single dimension, is introduced
Algorithm, goes to judge the song that user likes, such as bean cotyledon；3rd epoch are to be recommended by social networks and data mining,
By introducing social graph and interest graph, user's music interested is excavated.
The content of the invention
Technical problem：Music recommends to recommend with text and other information recommendations make a big difference.One, music recommendation has
Suitable subjectivity, can be influenceed by many factors such as surrounding environment.Two, with available time, can be with broadcasting time
Changed with fashionable colors.Three, with ambiguity, music recommends often be accurate to certain head song, and compared to similar
The recommendation of song should more focus on the excavation of potential interest.Therefore, " special edition ", " song only are utilized with basic data digging method
The hardness classification of the label such as hand ", " style " recommends or going out suggested design according to historical record association mining can not all expire well
Foot is actually needed.
Technical scheme：For blindness and mechanical problem that traditional music is recommended, the present invention relates to a kind of new
Individualized music suggested design.Using data mining technology, have levels by subscriber segmentation, according to the historical record of user, point
Separate out a variety of demands.Be combined with social networks simultaneously, the music liked according to user and disagreeable music to
Family is clustered, and to each user generation interest tendency label, recommended user adds as a friend neighboring user.Then with good friend
It is data set to enclose, and is recommended using correlation rule so that recommended more accurate efficient.
The operational process of one music commending system for taking this programme is as follows：
1. a kind of music based on social networks recommends method, it is characterised in that the recommendation method and step is as follows：
Step 1. points out User logs in, and the historical record of the user is inquired about in background data base；If the user is
Then execution step 2~step 4 is once logged in, step 5~step 7, step 8 of finally seeking unity of action otherwise is performed；
Music number of times sequence of the step 2. by music commending system in current Qu Ku, selects most popular song
Recommend as initialization playlist；
Step 3. commending system while waiting user to listen to, point out user to be liked every first song, disliked and
Skip the selection of option and classification analysis is carried out to these evaluation results, calculateWherein, i represents certain
Individual user, n represents the sum for liking song, and m represents the sum of disagreeable song, ri,kRepresent whether user i likes kth song,
ti,kRepresent the whether disagreeable kth songs of user i；
Simultaneously, the selection result that will be collected into step 3 carries out cluster analysis to step 4., calculates any two user i, j
Between similarity sim (i,j), adjacent friend circle is found, point out the interest tendency label of neighboring user, favorite song, the letter of special edition
Breath, recommended user is added to good friend；Then step 8 is performed；
The user that step 5. was once logged in has storage of history data P in database, and commending system is from background data base
The buddy list and friend circle for reading the user play record；
The history that step 6. reads the user by commending system from background data base again plays record, if listen in the recent period
Frequent degree exceedes three times a day, every time more than half an hour, then performs step 7, otherwise skips step 7, directly performs step 8；
Step 7. commending system is the music of newest addition in user's commending friends circle；
Step 8. commending system is associated regular recommendation according to broadcasting record in the friend circle of the user, for the user pushes away
Music is recommended, and continues to record its evaluation result；
Step 9. can always monitor the interest migration factor ε=p of the user during music commending system runs2+d×
L, wherein p represent that the unit interval skips the ratio of song and total recommendation song, and d represents that deleting disagreeable song accounts for all recommendation songs
Bent ratio, l represent the deleted song once liked account for it is all like the ratio of song, if meeting default threshold 0.6,
Step 10~step 12 is performed, step 13 is otherwise directly performed；
Step 10. re-starts classification analysis to user, updates the interest tendency label of user；
Step 11. re-starts cluster analysis to user, updates friend recommendation list, adds new good friend, deletes old friendship friend；
Step 12. is associated regular recommendation according to broadcasting record in the friend circle after renewal；
Step 13. continues to provide the user music service until user exits the service.
Beneficial effect：The present invention proposes a kind of new individualized music suggested design, and the main advantage of the program is：
First, user's request has been segmented, it is to avoid unexpected winner song cannot be introduced into the defect of recommendation list；
Two are combined with social networks, reduce the scope of data mining, it is to avoid blindness that traditional music is recommended and
Mechanicalness so that recommend more accurate efficient；
3rd, the concept of interest migration factor is proposed, the ability of machine learning is increased, the user that more fits needs, and chases after
Track changes in demand, and data mining error can be corrected, it is to avoid repeat to recommend；
4th, with cloud computing technology, to process the music information and user data of magnanimity, with good expansion.
Brief description of the drawings
Fig. 1 music recommended flowsheet figures.
Fig. 2 customer relationship logic charts.
Fig. 3 user interest transition graphs.
Specific embodiment
Individualized music recommend it is critical that each user is an individual, only got the spy of each individuality clear
Levy, can just clear the contact between individuality, efficiently recommended.Therefore, user account is using as distinguishing individual unique mark
Know.
If logging in for the first time, then 20 most popular songs are selected as initial list, be used to collect the first of user
Step information, if the simply very short visitor of listening period, most hot song is also sufficient for demand.
Interest is inclined to label：User can select to like when listening to, and dislike, the option such as skip, and these selection results are entered
Row classification, the music that will be liked and disagreeable music are classified respectively, according to music attribute in itself, such as style of song, special edition, singer
Deng being user's generation interest tendency label, including singer is liked, music style at most plays song, most likes song etc..Example
Such as：(user A, rock and roll, May, favorite song《It is stubborn》)
With user as an individual, carrying out the similarity between cluster analysis, user according to these results is
Wherein, i, j represent two users, and n represents the sum for liking song, and m represents the sum of disagreeable song, ri,k,rj,k
Represent whether user i, j like kth song, ti,k,tj,kRepresent user i, the whether disagreeable kth songs of j.
It is that user sorts out according to the similarity of user and each friend circle central point, finds adjacent friend circle, points out adjacent use
The information such as the interest tendency label at family, favorite song, special edition, recommended user is added to good friend, is by user oneself judgement finally
No addition, it is to avoid systematic error.
Different users is different for the demand recommended, and most notably, it is tired that repetition listens a small amount of concert to produce
Labor sense, and good commending system should be able to digging user potential interest.Therefore, only it is merely constantly to repeat to recommend
It is inadequate that user likes the hot music of type.If the user listens to very frequently, should just recommend the song of newest issue
And its potential interest is excavated in circle of friends compared with the song of unexpected winner.
Regular recommendation is associated according to record is played in friend circle, rule is searched according to itself historical record, pushed away
Recommend result.Data set is reduced within friend circle, that is, reduces amount of calculation, facilitate decimation rule, the essence of recommendation is improve again
True property.
Interest migration factor：The hobby of user be not it is unalterable, can over time with the factor such as environment and change,
And it is just very accurate also to be started by the result obtained by data mining.So occurring that continuous recommended user begs for
Situations such as music detested, the music frequently skipped, and user delete the music once liked.Therefore, interest is added to move
Move the concept of factor ε to learn the hobby of user, the error that the change and equilibrium criterion for reacting user interest are excavated.
ε=p2+d×l
Wherein, p represents that the unit interval skips the ratio of song and total recommendation song, and d represents that deleting disagreeable song accounts for institute
There is the ratio for recommending song, l represents that the deleted song once liked accounts for all ratios for liking song.
When ε meets certain threshold values, it is necessary to which proposed standard is modified, classification analysis is re-started to user, more
The interest tendency label of new user, cluster analysis is re-started to user, updates friend recommendation list, adds new good friend, is deleted
Old friendship friend.The determination of specific threshold values needs many practical factors such as consideration library sum, total number of users, average line duration,
A fixed value cannot simply be determined, should at any time be adjusted according to actual conditions, acquiescence is set to 0.6.
Can be commented on and visible to the user in friend circle per first song, nearest part be listed file names with during recommendation and is commented
By, increase the surcharge of music, improve stickiness.
Claims (1)
1. a kind of music based on social networks recommends method, it is characterised in that the recommendation method and step is as follows：
Step 1. points out User logs in, and the historical record of the user is inquired about in background data base；If the user is for the first time
Log in and then perform step 2~step 4, otherwise perform step 5~step 7, step 8 of finally seeking unity of action；
Music number of times sequence of the step 2. by music commending system in current Qu Ku, selects most popular song conduct
Initialization playlist is recommended；
Step 3. commending system points out user to like, disliked and skip every first song while waiting user to listen to
The selection of option simultaneously carries out classification analysis to these evaluation results, calculatesWherein, i represents that certain is used
Family, n represents the sum for liking song, and m represents the sum of disagreeable song, ri,kRepresent whether user i likes kth song, ti,k
Represent the whether disagreeable kth songs of user i；
Simultaneously, the selection result that will be collected into step 3 carries out cluster analysis to step 4., calculates between any two user i, j
Similarity sim (i,j), the adjacent friend circle of searching, the interest tendency label of prompting neighboring user, favorite song, the information of special edition,
Recommended user is added to good friend；Then step 8 is performed；
The user that step 5. was once logged in has storage of history data P in database, and commending system reads from background data base
The buddy list and friend circle of the user play record；
The history that step 6. reads the user by commending system from background data base again plays record, if that listens in the recent period is frequent
Degree exceedes three times a day, every time more than half an hour, then performs step 7, otherwise skips step 7, directly performs step 8；
Step 7. commending system is the music of newest addition in user's commending friends circle；
Step 8. commending system is associated regular recommendation according to broadcasting record in the friend circle of the user, is that the user recommends sound
It is happy, and continue to record its evaluation result；
Step 9. can always monitor the interest migration factor ε=p of the user during music commending system runs2+ d × l, wherein
P represents that the unit interval skips the ratio of song and total recommendation song, and d is represented and deleted the ratio that disagreeable song accounts for all recommendation songs
Example, l represent the deleted song once liked account for it is all like the ratio of song, if meeting default threshold 0.6, perform step
Rapid 10~step 12, otherwise directly performs step 13；
Step 10. re-starts classification analysis to user, updates the interest tendency label of user；
Step 11. re-starts cluster analysis to user, updates friend recommendation list, adds new good friend, deletes old friendship friend；
Step 12. is associated regular recommendation according to broadcasting record in the friend circle after renewal；
Step 13. continues to provide the user music service until user exits the service.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410192981.4A CN104008138B (en) | 2014-05-08 | 2014-05-08 | A kind of music based on social networks recommends method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410192981.4A CN104008138B (en) | 2014-05-08 | 2014-05-08 | A kind of music based on social networks recommends method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104008138A CN104008138A (en) | 2014-08-27 |
| CN104008138B true CN104008138B (en) | 2017-06-30 |
Family
ID=51368795
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410192981.4A Expired - Fee Related CN104008138B (en) | 2014-05-08 | 2014-05-08 | A kind of music based on social networks recommends method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104008138B (en) |
Families Citing this family (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105893361A (en) * | 2014-09-23 | 2016-08-24 | 江苏奥博洋信息技术有限公司 | A method for structured classification of users in socialized media |
| CN104408051B (en) * | 2014-10-28 | 2019-04-09 | 广州酷狗计算机科技有限公司 | Song recommendations method and device |
| CN104615749A (en) * | 2015-02-12 | 2015-05-13 | 深圳市欧珀通信软件有限公司 | Ring tone recommendation method and ring tone recommendation device |
| CN104899265B (en) * | 2015-05-21 | 2018-07-20 | 广东小天才科技有限公司 | Information recommendation method and system |
| WO2017124394A1 (en) * | 2016-01-21 | 2017-07-27 | 阮元 | Method for automatically recommending resources by vehicle-mounted computer and recommendation system |
| CN105975483B (en) * | 2016-04-25 | 2020-02-14 | 北京三快在线科技有限公司 | Message pushing method and platform based on user preference |
| CN106021302A (en) * | 2016-05-04 | 2016-10-12 | 北京思特奇信息技术股份有限公司 | Intelligent recommendation technique based wireless music recommendation method and system |
| 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 |
| CN106850417A (en) * | 2017-04-06 | 2017-06-13 | 北京深思数盾科技股份有限公司 | A kind of method and device for setting up user-association relation |
| CN107368552A (en) * | 2017-06-30 | 2017-11-21 | 广东欧珀移动通信有限公司 | A kind of friend recommendation method, apparatus, storage medium, server and terminal |
| CN110633408B (en) * | 2018-06-20 | 2024-03-15 | 北京正和岛信息科技有限公司 | Intelligent business information recommendation method and system |
| CN109299316B (en) * | 2018-11-09 | 2023-04-18 | 平安科技（深圳）有限公司 | Music recommendation method and device and computer equipment |
| CN109637559A (en) * | 2018-11-10 | 2019-04-16 | 东莞市华睿电子科技有限公司 | A kind of method for playing music applied to bullet train |
| CN110321478B (en) * | 2019-05-27 | 2024-11-08 | 腾讯科技（北京）有限公司 | Information recommendation method, device, equipment and medium |
| CN110297939A (en) * | 2019-06-21 | 2019-10-01 | 山东科技大学 | A kind of music personalization system of fusion user behavior and cultural metadata |
| CN110704744A (en) * | 2019-09-30 | 2020-01-17 | 北京金山安全软件有限公司 | Method and device for recommending target object to user and electronic equipment |
| CN110851651B (en) * | 2019-11-08 | 2022-07-22 | 杭州小影创新科技股份有限公司 | Personalized video recommendation method and system |
| CN111314205B (en) * | 2020-01-16 | 2023-06-27 | 广州酷狗计算机科技有限公司 | Instant messaging matching method, device, system, equipment and storage medium |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7206775B2 (en) * | 2000-07-06 | 2007-04-17 | Microsoft Corporation | System and methods for the automatic transmission of new, high affinity media |
-
2014
- 2014-05-08 CN CN201410192981.4A patent/CN104008138B/en not_active Expired - Fee Related
Non-Patent Citations (1)
| Title |
|---|
| Mixing It Up: Recommending Collections of Items;Derek L. Hansen;《Classifying and Recommending Content》;20090408;第5页第7章第1节第1段，第8页第8章第4节，表3 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104008138A (en) | 2014-08-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104008138B (en) | A kind of music based on social networks recommends method | |
| US8869042B2 (en) | Recommendation engine | |
| Hodkinson | Translocal Connections | |
| Laplante | Improving music recommender systems: What can we learn from research on music tastes? | |
| KR101870353B1 (en) | Proposing system for search advertising using issue keyword and associated keyword | |
| CN103455538B (en) | Information processing unit, information processing method and program | |
| US20210294843A1 (en) | Playlist preview | |
| US20140095307A1 (en) | Spotting trends by identifying influential consumers | |
| CN106846061A (en) | Potential user's method for digging and device | |
| CN101446950A (en) | Grouping and weighting media categories with time periods | |
| CN107885745A (en) | A kind of song recommendations method and device | |
| Millward et al. | A ‘different class’? Homophily and heterophily in the social class networks of Britpop | |
| US20220122147A1 (en) | Emotion calculation device, emotion calculation method, and program | |
| US9189802B2 (en) | Method for contents recommendation | |
| US20210149951A1 (en) | Audio and Visual Asset Matching Platform | |
| Guy et al. | Islands in the stream: A study of item recommendation within an enterprise social stream | |
| Ferwerda et al. | Exploring online music listening behaviors of musically sophisticated users | |
| Kudori | Event recommendation system using hybrid method based on mobile device | |
| Zhang et al. | Musical preference in an online music community in China | |
| Hamilton et al. | Rate and review: Exploring listener motivations for engagement with music podcasts | |
| US20210357445A1 (en) | Multimedia asset matching systems and methods | |
| Schaller et al. | Relating user interaction to experience during festivals | |
| Kim et al. | TiMers: time-based music recommendation system based on social network services analysis | |
| Saldana | Differences in How Boomers and Millennials Experience Listening to Recorded Music | |
| Dang et al. | FmFinder: Search and filter your favorite songs |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20170630 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |