CN106557560A - Level music based on user interest recommends method - Google Patents
Level music based on user interest recommends method Download PDFInfo
- Publication number
- CN106557560A CN106557560A CN201610995675.3A CN201610995675A CN106557560A CN 106557560 A CN106557560 A CN 106557560A CN 201610995675 A CN201610995675 A CN 201610995675A CN 106557560 A CN106557560 A CN 106557560A
- Authority
- CN
- China
- Prior art keywords
- user
- music
- interest
- list
- user interest
- 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
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
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)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The present invention provides a kind of level music based on user interest and recommends method, including step：User is read from log system and plays daily record；Analysis user plays daily record, obtains user behavior characteristic；Basic music list is obtained with reference to user behavior characteristic using basic algorithm；The interest model of user is built according to the user behavior characteristic；Reproducible history music list is obtained using the user interest model；User interest degree is obtained according to the user behavior characteristic, the music list that user may be interested is obtained according to the user interest degree；The basic music list, the reproducible history music list and the possible music list interested are combined into recommendation according to the user interest degree.Music recommendation can be carried out with reference to the potential interest of user using the present invention, improve the recommendation degree of accuracy that music recommends method, user interest fluctuation change can be reflected in real time, user's viscosity is improved.
Description
Technical field
The present invention relates to music data process field, more particularly to a kind of level music recommendation side based on user interest
Method and system.
Background technology
With the high speed development of internet industry in recent years, the rapid expansion of magnanimity music sources makes user for music
Selection becomes particularly difficult.Therefore, how quickly and efficiently suitable music is recommended to become music software to user rapid
The key subject of occuping market.Music recommends the recommendation for including based on music content, based on the recommendation of music relatedness, based on knowing
The various ways such as the recommendation of knowledge, collaborative filtering recommending, the technology for adopting at present mainly have：1st, when receiving at least one terminal
During music recommendation request, the music preferences of at least one terminal are obtained, and a terminal-pair answers at least one music preferences；According to each
The music preferences of individual terminal, generate the common music preferences of at least one terminal；It is according to common music preferences, whole at least one
End carries out music recommendation.2nd, the behavior of listening to of user is built on the basis of song to be modeled as some implicit theme probability distribution
Mould is multidimensional time-series, and then the behavioural habits of the method digging user by Multidimensional time series analysis, and final from time
Recommend suitable song for user in selecting song database.3rd, on the basis that song is modeled as some implicit theme probability distribution
On user listened to into behavior modeling for multidimensional time-series, and then the method digging user by Multidimensional time series analysis
Behavioural habits, and finally recommend suitable song for user from candidate song data base.But to the potential hobby of user
Lack and consider, it is impossible to reflection user interest fluctuation change in real time.
The content of the invention
Based on this, the purpose of the embodiment of the present invention be provide a kind of level music based on user interest recommend method and
System, which can carry out music recommendation with reference to the potential interest of user, improve the recommendation degree of accuracy that music recommends method, can be in real time
Reflection user interest fluctuation change, improves user's viscosity.
To reach above-mentioned purpose, the embodiment of the present invention is employed the following technical solutions：
A kind of level music based on user interest recommends method, including step：
S1, from log system read user play daily record；
S2, analysis user play daily record, obtain user behavior characteristic；Wherein, the user behavior characteristic includes what is listened to
Music name, the time for listening to the music, the duration for listening to the music, the number of times for listening to the music, the ID for listening to the music,
The type of the music, geographical location information, user's use habit information residing for user；
S3, basic music list is obtained with reference to user behavior characteristic using basic algorithm；
S4, the interest model that user is built according to the user behavior characteristic；
S5, reproducible history music list is obtained using the user interest model；
S6, according to the user behavior characteristic obtain user interest degree, according to the user interest degree obtain user may
Music list interested；
S7, according to the user interest degree by the basic music list, the reproducible history music list and
The possible music list interested is combined recommendation.
And a kind of level music commending system based on user interest, including：
Reading unit, plays daily record for reading user from log system；
Analytic unit, plays daily record for analyzing user, obtains user behavior characteristic；Wherein, the user behavior characteristic
Including the music name listened to, the time for listening to the music, the duration for listening to the music, the number of times for listening to the music, listen to the sound
Happy ID, the type of the music, geographical location information, user's use habit information residing for user；
First acquisition unit, for obtaining basic music list with reference to user behavior characteristic using basic algorithm；
Model construction unit, for the interest model of user is built according to the user behavior characteristic；
Second acquisition unit, for obtaining reproducible history music list using the user interest model；
3rd acquiring unit, for obtaining user interest degree according to the user behavior characteristic, according to the user interest
Degree obtains the music list that user may be interested；
First recommendation unit, for according to the user interest degree by the basic music list, described reproducible go through
History music list and the possible music list interested are combined recommendation.
Using the present invention program, using user behavior characteristic, according to different approach obtain respectively basic music list, can
The history music list of reproduction and user may be interested music list, and obtain the interest-degree of user and in this, as sentencing
Other user enlivens the foundation of type；The user of type is enlivened for difference, basic music list, reproducible history music are arranged
The ratio that table and the possible music list interested of user are finally strategically distributed is combined recommendation, can be with reference to user
Potential interest carries out music recommendation, improves the recommendation degree of accuracy that music recommends method, can reflect that user interest fluctuation becomes in real time
Change, improve user's viscosity.
Description of the drawings
Accompanying drawing is, for providing a further understanding of the present invention, and to constitute the part of description, concrete with following
Embodiment is used for explaining the present invention together, but should not be construed as limiting the invention.In the accompanying drawings,
Fig. 1 is the schematic flow sheet that the level music based on user interest proposed by the present invention recommends method；
Fig. 2 shows the level music recommendation system framework figure based on user interest proposed by the present invention；
Fig. 3 is a structural representation of the level music commending system based on user interest proposed by the present invention；
Fig. 4 is the structural representation of first acquisition unit proposed by the present invention；
Fig. 5 is the structural representation of model construction unit proposed by the present invention；
Fig. 6 is the structural representation of second acquisition unit proposed by the present invention；
Fig. 7 is the structural representation of the 3rd acquiring unit proposed by the present invention.
Specific embodiment
To make the objects, technical solutions and advantages of the present invention become more apparent, below in conjunction with drawings and Examples, to this
Invention is described in further detail.It should be appreciated that specific embodiment described herein is only to explain the present invention,
Protection scope of the present invention is not limited.
Fig. 1 shows the flow chart that a kind of level music based on user interest recommends method, including step：
S1, from log system read user play daily record；
Specifically, the music log system from music servers reads user and plays daily record.
S2, analysis user play daily record, obtain user behavior characteristic；Wherein, the user behavior characteristic includes what is listened to
Music name, the time for listening to the music, the duration for listening to the music, the number of times for listening to the music, the ID for listening to the music,
The type of the music, geographical location information, user's use habit information residing for user；Wherein, the musical designation that user listens to refers to
Be music name, such as Li Rong is great《Model》；The time for listening to the music is the moment for listening to the music, geographical residing for user
Positional information refers to that user listens to current music position, and with regard to this position, we can be obtained by the network address, example
As identifying user is in south China, the north, western part or east, analysis geographical location information is mainly based upon different geographical
User may be differed in the preference type of music, and the user in the such as north may like bolder and more unconstrained, southern user's possibility
Like the tenderness music of some.User's use habit information refers to user and carries out the period preference that music is listened to, such as morning,
Noon, afternoon, evening etc..Personal like's information refers to the type of user's music interested, such as country music, Chinese music
Deng.
S3, basic music list is obtained with reference to user behavior characteristic using basic algorithm；
Specifically, the step includes：
S31, according to the user behavior characteristic construct ' user-music-label ' three-dimensional matrice；
S32, the three-dimensional matrice for launching each user, use ' user-label ' matrix to open up ' user-music ' matrix
Exhibition；
S33, recommendation is calculated based on user and based on the collaborative filtering blending algorithm of music using fusion to each user
List；
S34, the merging recommendation list simultaneously sort, and obtain the basic music list.
Step S33 is described more fully below, " fusion is used to each user based on user and the collaborative filtering based on music
Blending algorithm constructs recommendation list " process：
1) matrix after being expanded to user-music matrix according to user-label matrix is calculating and specific user's phase
Like the N number of user characteristicses of degree highest Top, the contribution based on User Part is calculated by following equation 1；
NuFor the neighbours of user u, Ov,iRepresent that user v evaluated music i that still user u was not in contact with.
2) matrix after being expanded to ' user-music ' matrix according to ' music-label ' matrix, using based on music class
Other TopN algorithms obtain recommending music Ni, calculate the contribution based on musical portions by formula 2；
NiFor the neighbours of music i, Ou,jMusic j that user u was evaluated is represented, w (i, j) represents the phase of article i and article j
Like degree.
3) fusion calculated based on user and based on music by following formula is scored：
4) basisScoring highest P song is chosen as recommendation results.
S4, the interest model that user is built according to the user behavior characteristic；
Specifically, step S4 includes step：
S41, the data set D1 comprising playback features and music categories label is constructed according to the user behavior characteristic with
And the data set D2 comprising playback features；Wherein described playback features include user to the broadcasting time of music, user to music
Average audition duration to music of total audition duration, user；
S42, using the sorting algorithm of machine learning, using the data set D1 as sorting algorithm input sample collection, instruction
The user interest model comprising feature weight is got, and stable user interest model is obtained after multiple repairing weld is trained.
S5, reproducible history music list is obtained using the user interest model；
Specifically, including step：
S51, using the data set D2 as the input of the stable user interest model, obtain owning in data set D2
Broadcasting probability P of the user to different music；
S52, according to the broadcasting probability P, probit highest N song is taken out to each user and obtains each user's
The reproducible history music list.
S6, according to the user behavior characteristic obtain user interest degree, according to the user interest degree obtain user may
Music list interested；
Specifically, including step：
S61, according to user behavior property calculation user interest angle value；Specifically, can be according in user behavior characteristic
' duration for listening to music, the number of times for listening to music, the type of music ' is referred to, user arranging different calculating weights
The type of the number of times+k3 × music of the duration+k2 of interest level=k1 × listen to music × listen to music, wherein, the weight of k3
More than the weight of k1, k2.
S62, the user interest angle value is carried out branch mailbox to carry out sliding-model control；
User interest angle value after S63, the discretization obtained according to step S62 recommends music list that may be interested,
The user low for user interest angle value, selects sound of the music number of times less than threshold value in user's music categories interested
Pleasure is recommended；The user high for user interest degree, recommends the music of the classification that the user do not listened.Wherein, user interest
That what is spent is high or low, can arrange corresponding quantitative criteria according to actual needs, when such as interest-degree is higher than 10, is that interest-degree is high.
When interest-degree is less than 10, it is that interest-degree is low.
S7, according to the user interest degree by the basic music list, the reproducible history music list and
The possible music list interested is combined recommendation.
Specifically, the step includes：The basic music list, described reproducible is preset to different user interest degrees
The recommendation ratio of history music list and the music list that may be interested, carries out music according to the recommendation ratio and pushes away
Recommend.
Refer to, various levels (basic music, reproducible history sound are pre-set for different user interest-degree
Happy, music that may be interested) music recommendation ratio, such as interest-degree D1 when, based on the recommendation ratio of the music of each level
Music：Reproducible history music：Energy music=5 interested：4：1；During interest-degree D2, the recommendation ratio of the music of each level
Based on music：Reproducible history music：Energy music=7 interested：2：1；During interest-degree D3, the music of each level is pushed away
Recommend music based on ratio：Reproducible history music：Energy music=4 interested：4：2；Etc..
Using the present invention program, using user behavior characteristic, according to different approach obtain respectively basic music list, can
The history music list of reproduction and user may be interested music list, and obtain the interest-degree of user and in this, as sentencing
Other user enlivens the foundation of type；The user of type is enlivened for difference, basic music list, reproducible history music are arranged
The ratio that table and the possible music list interested of user are finally strategically distributed is combined recommendation, can be with reference to user
Potential interest carries out music recommendation, improves the recommendation degree of accuracy that music recommends method, can reflect that user interest fluctuation becomes in real time
Change, improve user's viscosity.
Fig. 2 shows music recommendation system framework figure proposed by the present invention, reads from music commending system backstage first and uses
Daily record is played as core data in family, then carries out pretreatment to core data, obtains user behavior characteristic, then carries out model
Build, and user interest degree is obtained；In commending system layer, basic music row are obtained with reference to user behavior characteristic using basic algorithm
Table；Reproducible history music list is obtained using the user interest model for building；And used according to the user interest degree
The possible music list interested in family；It is last according to the user interest degree by the basic music list, described reproducible
History music list and the possible music list interested are combined recommendation, and result is pushed to service application layer.
Based on Spark, (Spark is class Hadoop increased income by UC Berkeley AMP lab to the present invention program
The general parallel computation frame of MapReduce, the Distributed Calculation that Spark is realized based on map reduce algorithms) big number
According to processing platform, distributed treatment mass users data are carried out using Spark, with enhanced scalability and higher process effect
Rate.
Correspondence said method, the invention allows for a kind of level music commending system based on user interest, its structure
Schematic diagram is Fig. 3, including：
Reading unit, plays daily record for reading user from log system；
Analytic unit, plays daily record for analyzing user, obtains user behavior characteristic；Wherein, the user behavior characteristic
Including the music name listened to, the time for listening to the music, the duration for listening to the music, the number of times for listening to the music, listen to the sound
Happy ID, the type of the music, geographical location information, user's use habit information residing for user；
First acquisition unit, for obtaining basic music list with reference to user behavior characteristic using basic algorithm；
Model construction unit, for the interest model of user is built according to the user behavior characteristic；
Second acquisition unit, for obtaining reproducible history music list using the user interest model；
3rd acquiring unit, for obtaining user interest degree according to the user behavior characteristic, according to the user interest
Degree obtains the music list that user may be interested；
First recommendation unit, for according to the user interest degree by the basic music list, described reproducible go through
History music list and the possible music list interested are combined recommendation.
Wherein, as shown in figure 4, the first acquisition unit includes：
Matrix construction unit, for constructing ' user-music-label ' three-dimensional matrice according to the user behavior characteristic；
Unit is expanded, and for launching the three-dimensional matrice of each user, ' user-label ' matrix is used to ' user-music '
Matrix is expanded；
Second recommendation unit, for being calculated based on user and the fusion of the collaborative filtering based on music using fusion to each user
Method constructs recommendation list；Merge the recommendation list and sort, obtain the basic music list.
Wherein, as shown in figure 5, the model construction unit includes：
Data set construction unit, for being constructed comprising playback features and music categories mark according to the user behavior characteristic
The data set D1 of the label and data set D2 comprising playback features；Wherein described playback features include broadcasting time of the user to music
Number, average audition duration of the user to total audition duration, user of music to music；
Model training unit, for the sorting algorithm using machine learning, using the data set D1 as sorting algorithm
Input sample collection, training obtain the user interest model comprising feature weight, after multiple repairing weld is trained obtain stable use
Family interest model.
Wherein, as shown in fig. 6, the second acquisition unit includes：
Probability calculation unit, obtains as the input of the stable user interest model for using the data set D2
Broadcasting probability P of all users to different music in data set D2；
4th acquiring unit, for according to the broadcasting probability P, taking out probit highest N song to each user
Obtain the described reproducible history music list of each user.
Wherein, as shown in fig. 7, the 3rd acquiring unit includes：
Interest-degree computing unit, for according to user behavior property calculation user interest angle value；
Discrete processes unit, for, the user interest angle value carried out branch mailbox to carry out sliding-model control；
3rd recommendation unit, for recommending possible music list interested according to the user interest angle value after discretization,
The user low for user interest angle value, selects sound of the music number of times less than threshold value in user's music categories interested
Pleasure is recommended；The user higher for user interest degree, recommends the music of the classification that the user do not listened.
Wherein, first recommendation unit is preset the basic music list, described is answered to different user interest degrees
The recommendation ratio of existing history music list and the music list that may be interested, carries out sound according to the recommendation ratio
It is happy to recommend.
Said system is embodied as detailed rules and regulations, and with reference to the description of corresponding method, here is omitted.
Using the present invention program, using user behavior characteristic, according to different approach obtain respectively basic music list, can
The history music list of reproduction and user may be interested music list, and obtain the interest-degree of user and in this, as sentencing
Other user enlivens the foundation of type；The user of type is enlivened for difference, basic music list, reproducible history music are arranged
The ratio that table and the possible music list interested of user are finally strategically distributed is combined recommendation, can be with reference to user
Potential interest carries out music recommendation, improves the recommendation degree of accuracy that music recommends method, can reflect that user interest fluctuation becomes in real time
Change, improve user's viscosity.
Without departing from the thought of the invention, combination in any is carried out to the various different embodiments of the present invention, all should
When being considered as present disclosure；In the range of the technology design of the present invention, various simple modifications are carried out to technical scheme
And the combination in any of the thought without prejudice to the invention that different embodiments are carried out, all should protection scope of the present invention it
It is interior.
Claims (12)
1. a kind of level music based on user interest recommends method, it is characterised in that including step：
S1, from log system read user play daily record；
S2, analysis user play daily record, obtain user behavior characteristic；Wherein, the user behavior characteristic includes the music listened to
Name, the time for listening to the music, the duration for listening to the music, the number of times for listening to the music, the ID for listening to the music, the sound
Happy type, geographical location information, user's use habit information residing for user；
S3, basic music list is obtained with reference to user behavior characteristic using basic algorithm；
S4, the interest model that user is built according to the user behavior characteristic；
S5, reproducible history music list is obtained using the user interest model；
S6, according to the user behavior characteristic obtain user interest degree, according to the user interest degree obtain user may feel emerging
The music list of interest；
S7, according to the user interest degree by the basic music list, the reproducible history music list and described
Possible music list interested is combined recommendation.
2. the level music based on user interest according to claim 1 recommends method, it is characterised in that step S3
Including step：
S31, according to the user behavior characteristic construct ' user-music-label ' three-dimensional matrice；
S32, the three-dimensional matrice for launching each user, use ' user-label ' matrix to expand ' user-music ' matrix；
S33, recommendation list is constructed based on user and based on the collaborative filtering blending algorithm of music using fusion to each user；
S34, the merging recommendation list simultaneously sort, and obtain the basic music list.
3. the level music based on user interest according to claim 1 recommends method, it is characterised in that step S4
Including step：
S41, the data set D1 comprising playback features and music categories label and bag are constructed according to the user behavior characteristic
Data set D2 containing playback features；Wherein described playback features include user to the broadcasting time of music, user to the total of music
The average audition duration of audition duration, user to music；
S42, using the sorting algorithm of machine learning, using the data set D1 as the input sample collection of sorting algorithm, train
To the user interest model comprising feature weight, stable user interest model is obtained after multiple repairing weld is trained.
4. the level music based on user interest according to claim 3 recommends method, it is characterised in that step S5
Including step：
S51, using the data set D2 as the input of the stable user interest model, obtain all users in data set D2
Broadcasting probability P to different music；
S52, according to the broadcasting probability P, probit highest N song is taken out to each user and is obtained described in each user
Reproducible history music list.
5. the level music based on user interest according to claim 1 recommends method, it is characterised in that step S6
Including step：
S61, according to user behavior property calculation user interest angle value；
S62, the user interest angle value is carried out branch mailbox to carry out sliding-model control；
User interest angle value after S63, the discretization obtained according to step S62 recommends music list that may be interested, for
The low user of user interest angle value, selects music number of times to enter less than the music of threshold value in user's music categories interested
Row is recommended；The user high for user interest degree, recommends the music of the classification that the user do not listened.
6. the level music based on user interest according to claim 1 recommends method, it is characterised in that step S7
Including step：
Different user interest degrees is preset the basic music list, the reproducible history music list and it is described can
The recommendation ratio of music list that can be interested, carries out music recommendation according to the recommendation ratio.
7. a kind of level music commending system based on user interest, it is characterised in that include：
Reading unit, plays daily record for reading user from log system；
Analytic unit, plays daily record for analyzing user, obtains user behavior characteristic；Wherein, the user behavior characteristic includes
The music name listened to, the time for listening to the music, the duration for listening to the music, the number of times for listening to the music, listen to the music
ID, the type of the music, geographical location information, user's use habit information residing for user；
First acquisition unit, for obtaining basic music list with reference to user behavior characteristic using basic algorithm；
Model construction unit, for the interest model of user is built according to the user behavior characteristic；
Second acquisition unit, for obtaining reproducible history music list using the user interest model；
3rd acquiring unit, for obtaining user interest degree according to the user behavior characteristic, obtains according to the user interest degree
Obtain the music list that user may be interested；
First recommendation unit, for according to the user interest degree by the basic music list, the reproducible history sound
Happy list and the possible music list interested are combined recommendation.
8. the level music commending system based on user interest according to claim 7, it is characterised in that described first obtains
Taking unit includes：
Matrix construction unit, for constructing ' user-music-label ' three-dimensional matrice according to the user behavior characteristic；
Unit is expanded, and for launching the three-dimensional matrice of each user, ' user-label ' matrix is used to ' user-music ' matrix
Expanded；
Second recommendation unit, for fusion is used to each user based on user and the collaborative filtering blending algorithm structure based on music
Produce recommendation list；Merge the recommendation list and sort, obtain the basic music list.
9. the level music commending system based on user interest according to claim 7, it is characterised in that the model structure
Building unit includes：
Data set construction unit, for being constructed comprising playback features and music categories label according to the user behavior characteristic
The data set D1 and data set D2 comprising playback features；Wherein described playback features include user to the broadcasting time of music,
Average audition duration of the user to total audition duration, user of music to music；
Model training unit, for the sorting algorithm using machine learning, using the data set D1 as sorting algorithm input
Sample set, training obtain the user interest model comprising feature weight, obtain stable user emerging after multiple repairing weld is trained
Interesting model.
10. the level music commending system based on user interest according to claim 9, it is characterised in that described second
Acquiring unit includes：
Probability calculation unit, obtains data for using the data set D2 as the input of the stable user interest model
Broadcasting probability P of all users to different music in collection D2；
4th acquiring unit, for according to the broadcasting probability P, taking out probit highest N song to each user and obtaining
The described reproducible history music list of each user.
The 11. level music commending systems based on user interest according to claim 7, it is characterised in that the described 3rd
Acquiring unit includes：
Interest-degree computing unit, for according to user behavior property calculation user interest angle value；
Discrete processes unit, for, the user interest angle value carried out branch mailbox to carry out sliding-model control；
3rd recommendation unit, for recommending possible music list interested according to the user interest angle value after discretization, for
The low user of user interest angle value, selects music number of times to enter less than the music of threshold value in user's music categories interested
Row is recommended；The user high for user interest degree, recommends the music of the classification that the user do not listened.
The 12. level music commending systems based on user interest according to claim 7, it is characterised in that described first
Recommendation unit presets the basic music list, the reproducible history music list and institute to different user interest degrees
The recommendation ratio of music list that may be interested is stated, and music recommendation is carried out according to the recommendation ratio.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610995675.3A CN106557560A (en) | 2016-11-11 | 2016-11-11 | Level music based on user interest recommends method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610995675.3A CN106557560A (en) | 2016-11-11 | 2016-11-11 | Level music based on user interest recommends method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106557560A true CN106557560A (en) | 2017-04-05 |
Family
ID=58444123
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610995675.3A Pending CN106557560A (en) | 2016-11-11 | 2016-11-11 | Level music based on user interest recommends method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106557560A (en) |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107862085A (en) * | 2017-11-30 | 2018-03-30 | 郑州云海信息技术有限公司 | A kind of music supplying system based on big data |
| CN107911428A (en) * | 2017-11-03 | 2018-04-13 | 维沃移动通信有限公司 | A kind of song method for pushing, device and mobile terminal |
| CN108766474A (en) * | 2018-06-04 | 2018-11-06 | 深圳市沃特沃德股份有限公司 | Vehicle-mounted music playback method and mobile unit |
| CN109447763A (en) * | 2018-11-12 | 2019-03-08 | 百度在线网络技术（北京）有限公司 | Method, apparatus, storage medium and the terminal device that musical works is recommended |
| CN109460486A (en) * | 2018-09-12 | 2019-03-12 | 珠海豆饭科技有限公司 | A kind of voice-based information recommendation method and system |
| CN110120845A (en) * | 2018-02-07 | 2019-08-13 | 上海博泰悦臻网络技术服务有限公司 | Radio station playback method and cloud server |
| CN110750455A (en) * | 2019-10-18 | 2020-02-04 | 北京大学 | Intelligent online self-updating fault diagnosis method and system based on system log analysis |
| CN110789471A (en) * | 2019-10-18 | 2020-02-14 | 中国第一汽车股份有限公司 | Vehicle-mounted music adjusting system and vehicle-mounted music adjusting method |
| CN111259191A (en) * | 2020-01-16 | 2020-06-09 | 石河子大学 | Music education learning system and method for primary and secondary schools |
| CN111460279A (en) * | 2020-02-25 | 2020-07-28 | 拉扎斯网络科技（上海）有限公司 | Information recommendation method and device, storage medium and computer equipment |
| CN111813918A (en) * | 2020-06-18 | 2020-10-23 | 国网上海市电力公司 | Scientific and technological resource recommendation processing method and device |
| CN112702465A (en) * | 2019-10-07 | 2021-04-23 | 东北林业大学 | Vehicle-mounted intelligent music playing app system based on android platform |
| CN112732971A (en) * | 2021-01-21 | 2021-04-30 | 广西师范大学 | Collaborative filtering music recommendation method based on labels |
| CN115134670A (en) * | 2022-07-22 | 2022-09-30 | 北京字跳网络技术有限公司 | Multimedia playing method, device, storage medium and program product |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105159973A (en) * | 2015-08-26 | 2015-12-16 | 努比亚技术有限公司 | Song recommendation apparatus and method |
| CN105550272A (en) * | 2015-12-09 | 2016-05-04 | 小米科技有限责任公司 | Song recommending method and device |
| CN105975472A (en) * | 2015-12-09 | 2016-09-28 | 乐视网信息技术（北京）股份有限公司 | Method and device for recommendation |
-
2016
- 2016-11-11 CN CN201610995675.3A patent/CN106557560A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105159973A (en) * | 2015-08-26 | 2015-12-16 | 努比亚技术有限公司 | Song recommendation apparatus and method |
| CN105550272A (en) * | 2015-12-09 | 2016-05-04 | 小米科技有限责任公司 | Song recommending method and device |
| CN105975472A (en) * | 2015-12-09 | 2016-09-28 | 乐视网信息技术（北京）股份有限公司 | Method and device for recommendation |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107911428A (en) * | 2017-11-03 | 2018-04-13 | 维沃移动通信有限公司 | A kind of song method for pushing, device and mobile terminal |
| CN107862085A (en) * | 2017-11-30 | 2018-03-30 | 郑州云海信息技术有限公司 | A kind of music supplying system based on big data |
| CN110120845A (en) * | 2018-02-07 | 2019-08-13 | 上海博泰悦臻网络技术服务有限公司 | Radio station playback method and cloud server |
| CN108766474A (en) * | 2018-06-04 | 2018-11-06 | 深圳市沃特沃德股份有限公司 | Vehicle-mounted music playback method and mobile unit |
| CN109460486A (en) * | 2018-09-12 | 2019-03-12 | 珠海豆饭科技有限公司 | A kind of voice-based information recommendation method and system |
| CN109447763A (en) * | 2018-11-12 | 2019-03-08 | 百度在线网络技术（北京）有限公司 | Method, apparatus, storage medium and the terminal device that musical works is recommended |
| CN112702465A (en) * | 2019-10-07 | 2021-04-23 | 东北林业大学 | Vehicle-mounted intelligent music playing app system based on android platform |
| CN110750455A (en) * | 2019-10-18 | 2020-02-04 | 北京大学 | Intelligent online self-updating fault diagnosis method and system based on system log analysis |
| CN110789471A (en) * | 2019-10-18 | 2020-02-14 | 中国第一汽车股份有限公司 | Vehicle-mounted music adjusting system and vehicle-mounted music adjusting method |
| CN111259191A (en) * | 2020-01-16 | 2020-06-09 | 石河子大学 | Music education learning system and method for primary and secondary schools |
| CN111460279A (en) * | 2020-02-25 | 2020-07-28 | 拉扎斯网络科技（上海）有限公司 | Information recommendation method and device, storage medium and computer equipment |
| CN111813918A (en) * | 2020-06-18 | 2020-10-23 | 国网上海市电力公司 | Scientific and technological resource recommendation processing method and device |
| CN111813918B (en) * | 2020-06-18 | 2024-06-14 | 国网上海市电力公司 | Scientific and technological resource recommendation processing method and device |
| CN112732971A (en) * | 2021-01-21 | 2021-04-30 | 广西师范大学 | Collaborative filtering music recommendation method based on labels |
| CN115134670A (en) * | 2022-07-22 | 2022-09-30 | 北京字跳网络技术有限公司 | Multimedia playing method, device, storage medium and program product |
| WO2024016902A1 (en) * | 2022-07-22 | 2024-01-25 | 北京字跳网络技术有限公司 | Multimedia playing method and device, storage medium, and program product |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106557560A (en) | Level music based on user interest recommends method | |
| CN105068661B (en) | Man-machine interaction method based on artificial intelligence and system | |
| CN108108821A (en) | Model training method and device | |
| CN110442790A (en) | Recommend method, apparatus, server and the storage medium of multi-medium data | |
| CN105808590B (en) | Search engine implementation method, searching method and device | |
| CN103870001B (en) | A kind of method and electronic device for generating candidates of input method | |
| CN106126524B (en) | Information pushing method and device | |
| CN106776928A (en) | Recommend method in position based on internal memory Computational frame, fusion social environment and space-time data | |
| CN103365936A (en) | Video recommendation system and method thereof | |
| CN103455538B (en) | Information processing unit, information processing method and program | |
| CN110134871B (en) | Dynamic course recommendation method based on course and learner network structure | |
| CN110110225A (en) | Online education recommended models and construction method based on user behavior data analysis | |
| Dubois et al. | The social conditions of cultural domination: Field, sub-field and local spaces of wind music in France | |
| CN111460145A (en) | Learning resource recommendation method, device and storage medium | |
| TW201820172A (en) | System, method and non-transitory computer readable storage medium for conversation analysis | |
| CN109615009A (en) | Learning content recommendation method and electronic equipment | |
| Piatetsky-Shapiro et al. | What are the grand challenges for data mining? KDD-2006 panel report | |
| Arai et al. | Predicting quality of answer in collaborative Q/A community | |
| CN110942376A (en) | Fusion method of real-time multi-recall strategy of audio products | |
| Li et al. | An adaptive course generation framework | |
| CN101655853A (en) | Device and method for building model | |
| CN108140034A (en) | Content item is selected using lexical item of the topic model based on reception | |
| CN113657766A (en) | Tourist attraction joy index metering method based on tourist multi-metadata | |
| Lionello et al. | Interactive exploration of musical space with parametric t-SNE | |
| CN118152667B (en) | Problem recommendation method for student preference |
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: 20170405 |
|
| RJ01 | Rejection of invention patent application after publication |