CN104731954A - Music recommendation method and system based on group perspective - Google Patents
Music recommendation method and system based on group perspective Download PDFInfo
- Publication number
- CN104731954A CN104731954A CN201510153358.2A CN201510153358A CN104731954A CN 104731954 A CN104731954 A CN 104731954A CN 201510153358 A CN201510153358 A CN 201510153358A CN 104731954 A CN104731954 A CN 104731954A
- Authority
- CN
- China
- Prior art keywords
- user
- song
- classification
- songs
- group
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- 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
-
- 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
- G06F16/9537—Spatial or temporal dependent retrieval, e.g. spatiotemporal queries
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)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention provides a music recommendation method and system based on group perspective. User characteristics and the favorite categories of songs listened to by users are analyzed according to user history logs, then user sets are divided according to the user characteristics and the favorite categories, user groups with the same general characteristics are built, and the user groups to which the users belong are analyzed; the collaborative filtering fusion algorithm is adopted to generate a recommendation song list, and the recommendation song list is pushed to the users. In the whole process, accurate user group division is carried out for a huge number of users, user interesting multidimensional perspective is achieved by deeply excavating the user data, and the music recommendation precision is improved. In addition, the user groups with the same general characteristics are built, and then the hybrid recommendation algorithm is used, so that music recommendation accuracy and reliability are further improved.
Description
Technical field
The present invention relates to music recommend technical field, particularly relate to and have an X-rayed music recommend method and system based on group.
Background technology
Along with the high speed development of internet industry in recent years, the rapid expansion of magnanimity music sources makes user become particularly difficulty for the selection of music.
Existing music recommend comprises the various ways such as the recommendation based on music content, the recommendation based on music relevance, Knowledge based engineering recommendation, collaborative filtering recommending.Although existing multiple different music recommend method at present, general music recommend method ubiquity recommends the problem that degree of accuracy is not high.
Summary of the invention
Based on this, be necessary to there is the problem of recommending degree of accuracy not high for general music recommend method, provide a kind of recommend degree of accuracy high have an X-rayed music recommend method and system based on group.
One has an X-rayed music recommend method based on group, comprises step:
Obtain user's history log, described user's history log record comprises user and listens to geographical location information residing for song title, user, user's use habit information, customer consumption ability information, individual subscriber preference information and age of user information;
According to described user's history log, analyze user in user personality and preset musical storehouse and listen to the hobby classification of song;
Listen to the hobby classification of song according to user in described user personality and preset musical storehouse, divide user's set, be built with the customer group of identical general character;
Analyze user's owning user group, adopt collaborative filtering blending algorithm to produce and recommend list of songs;
Push described recommendation list of songs.
One has an X-rayed music recommend system based on group, comprising:
Acquisition module, for obtaining user's history log, described user's history log record comprises user and listens to geographical location information residing for song title, user, user's use habit information, customer consumption ability information, individual subscriber preference information and age of user information;
Hobby category analysis module, for according to described user's history log, analyzes user in user personality and preset musical storehouse and listens to the hobby classification of song;
Customer group builds module, for listening to the hobby classification of song according to user in described user personality and preset musical storehouse, dividing user's set, being built with the customer group of identical general character;
List Generating Module, for analyzing user's owning user group, adopting collaborative filtering blending algorithm to produce and recommending list of songs;
Pushing module, for pushing described recommendation list of songs.
The present invention is based on group and have an X-rayed music recommend method and system, according to user's history log, analysis user personality and user listen to the hobby classification of song, again according to user personality and hobby classification, divide user's set, be built with the customer group of identical general character, analyze user's owning user group, adopt collaborative filtering blending algorithm to produce and recommend list of songs, push and recommend list of songs to user.In whole process, carry out customer group accurately for huge user to divide, by excavating the degree of depth of user data, realize having an X-rayed the various dimensions of user interest, improve music recommend precision, in addition, build the customer group with identical general character, combine mixing proposed algorithm again, further improve accuracy and the reliability of music recommend.
Accompanying drawing explanation
Fig. 1 the present invention is based on the schematic flow sheet that group has an X-rayed one of them embodiment of music recommend method;
Fig. 2 is the network diagram of user interest;
Fig. 3 the present invention is based on the structural representation that group has an X-rayed one of them embodiment of music recommend system.
Embodiment
The present invention is based on group and have an X-rayed the technical scheme of music recommend method and system mainly based on following theory:
1, based on transverse dimensions (song classification) customer group division methods Main Basis be music attribute label, i.e. music categories label, by the music interest common feature organizing user group of user.According to different labeling according to carrying out various dimensions segmentation.Such as, common song classification comprises " popular " and " rock and roll " (dividing according to style), " Japanese " and " Guangdong language " (dividing according to languages), " violin " and " piano " (dividing according to musical instrument), " Beethovan " and " Bach " (dividing according to artist) etc.The division of various dimensions customer group can be carried out to platform user by song class label, customer group inner analysis (namely there is in hobby higher similarity), reduce customer group scope, improve and recommend degree of accuracy.
2, based on longitudinal dimension (user personality) customer group division methods Main Basis be user tag, i.e. user property, by the common feature organizing user group of user personality.Wherein, user property, based on the analysis result of User action log, carries out various dimensions segmentation according to different labeling foundation.Such as, common attribute (user personality) label dimension comprises " teenager ", " teenager ", " youth ", " grow up " and " old age " (dividing according to age level), " country music " and " Chinese music " (dividing according to personal like), " noon " and " evening " (dividing according to use habit), " northeast ", " south China ", " northwest " and " East China " (dividing according to geographical location information), " by force " (monthly average consumption more than 1000 yuan), " stronger " (monthly average consumption more than 800 yuan) and " medium " (monthly average consumption more than 500 yuan) (dividing according to consuming capacity information) etc.In addition, by investigating the interest in music distribution of user's individuality, excavate the customer group (namely across the customer group of multiple transverse dimensions, its hobby the exists higher similarity) music of different dimensions label being had to inner link (mainly " liking ").Such as, the music that the user with " Beethovan " label music may generally can like with " piano " label is liked.This contact can not be embodied by the comparison of transverse dimensions, therefore need by excavating the investigation of longitudinal dimension, find the customer group across the interest of multiple transverse dimensions, contribute to the inner link of digging user point of interest, improve the accuracy of commending system.
As shown in Figure 1, one has an X-rayed music recommend method based on group, comprises step:
S100: obtain user's history log, described user's history log record comprises user and listens to geographical location information residing for song title, user, user's use habit information, customer consumption ability information, individual subscriber preference information and age of user information.
User's history log can obtain from the data of historical record.User's history log record includes but not limited to that user listens to geographical location information residing for song title, user, user's use habit information, customer consumption ability information, individual subscriber preference information and age of user information.User listens to song title and refers to song names, the such as May " content with one's lot ", geographical location information residing for user refers to that user listens to current song position, about this position, we can obtain a general position by positioning system, such as identify that user is in south China, the north, western part or east, analyzing geographical location information mainly may be not identical in the preference type of song based on the user of different geographical, user as the north may like bolder and more unconstrained, and southern user may like some song of tenderness.User's use habit information refers to user and carries out the period preference that music listens to, as morning, noon, afternoon, evening etc.Consuming capacity information refers to the information of the consuming capacity power being calculated user by user's consumer behavior in the past, as strong (monthly average consumption more than 1000 yuan), comparatively strong (monthly average consumption more than 800 yuan), medium (monthly average consumption more than 500 yuan) etc.Personal like's information refers to the type of the interested music of user, as country music, Chinese music etc.Age level information refers to the age place age level of user, as teenager, teenager, youth, adult, old etc.
S200: according to described user's history log, analyzes user in user personality and preset musical storehouse and listens to the hobby classification of song.
Preset musical storehouse is the database preset, and comprising there being a large amount of songs, user can search in this music libraries, audition, download the song oneself liked.Based on the user's history log obtained, we can analyze user personality from multiple angle (geography, custom, consuming capacity, preference information and age information etc.), and can listen to song based on user and analyze the hobby classification that user listens to song.
Wherein in an embodiment, step S200 specifically comprises:
Step one: the song in preset musical storehouse is classified, and the title of song and quantity under recording every kind.
Here classify according to can be that in style as the aforementioned, languages, musical instrument or artist, wherein a kind of mode is classified, record title and the quantity of song under every kind after song is classified.Such as classify according to style, be divided into nationality, rock and roll and popular, we record this three kind, and under recording national classification, number of songs is X, and under category Rock, number of songs is Y, and under popular classification, number of songs is Z.
Step 2: according to user's history log, analyzes user personality and obtains user and listen to song, analyzes user and listens to song generic.
After acquisition user listens to song, according to the principle of classification in above-mentioned steps, user can be analyzed and listen to song generic.
Step 3: calculate user and listen to the ratio that single classification number of songs and user listened to all number of songs, calculate the ratio of all number of songs in single classification number of songs and preset musical storehouse.
Be similar to TF-IDF (term frequency – inverse document frequency) method of weighting, for the main thought of label weighting is: if the frequency TF occurred in the song listened to user of certain label is high, and occur fewer in other songs, then think that this label has good class discrimination ability, be applicable to for classification.Similar, to a designated user, TF refers to the frequency occurred in the label of all songs that some given labels are listened to this user.A certain specific music label is for a certain specific user, its importance can be expressed as: molecule is all music numbers comprising this label (song classification) that user listened to, and denominator is then the such mark of all music numbers listened to user.IDF is the tolerance of a music label general importance.The IDF of a certain particular words, can by all music numbers of this label divided by all music numbers of music libraries, then the business obtained are taken the logarithm and obtain.
Step 4: listen to according to user the ratio that single classification number of songs and user listened to all number of songs in the ratio of all number of songs and single classification number of songs and preset musical storehouse, analyze user in preset musical storehouse and listen to the hobby classification of song.
Non-essential, user can be listened to the ratio that ratio that single classification number of songs and user listened to all number of songs is multiplied by all number of songs in single classification number of songs and preset musical storehouse and obtain a result of calculation.This result of calculation directly can characterize user in preset musical storehouse and listen to the hobby of song.Namely this result of calculation can as a weighted value of user preferences song classification.At some time, possibility song categorical measure is many, for reducing data processing amount, we can set a default weight threshold, the song classification lower than default weight threshold are deleted, focus in a classification by unified for the song in these classifications, such as other class.
S300: listen to the hobby classification of song according to user in described user personality and preset musical storehouse, divides user's set, is built with the customer group of identical general character.
After acquisition user personality and user listen to the hobby classification of song, will associate between the two, and divide user's set, be built with the customer group of identical general character.
Wherein in an embodiment, step S300 specifically comprises:
Step one: according to user personality, once divides user's set, obtains preliminary user set.
As described above, user personality is based on user's history log, and obtain from multiple angle (geography, custom, consuming capacity, preference information and age information etc.) analysis, user personality tentatively can distinguish user, obtains preliminary user set.
Step 2: the hobby classification listening to song according to user in preset musical storehouse, carries out fuzzy matching to described preliminary user set, obtains similar users set.
The hobby classification listening to song according to user carries out fuzzy matching, carries out secondary set division, obtains similar users set.So far, hobby classification two aspects having listened to song according to user in user personality and preset musical storehouse respectively carry out preliminary arrangement, division to user's set.
Step 3: according to user personality, carries out Similarity Measure to similar users set, builds user network.
Step 4: adopt user network described in the community discovery Algorithm Analysis in complex network, be built with the customer group of identical general character.
The hobby classification listening to song according to user in user personality and preset musical storehouse divides user, user can be divided into different sets, by building network and using the method for complex network-community discovery to find out the customer group with identical general character.
S400: analyze user's owning user group, adopts collaborative filtering blending algorithm to produce and recommends list of songs.
Use collaborative filtering blending algorithm to produce according to user's owning user group and recommend list of songs.Step S300 is obtained to the customer group structuring user's-music-song classification three-dimensional matrice of identical general character, use user-song classification matrix and article-music-song classification matrix, the augmented matrix of formation calculates recommendation list.
S500: push described recommendation list of songs.
After generating recommendations list of songs, can adopt various ways that list of songs is pushed to user.Such as directly be presented at the display interface of user's playback equipment.
The present invention is based on group and have an X-rayed music recommend method, according to user's history log, analysis user personality and user listen to the hobby classification of song, again according to user personality and hobby classification, divide user's set, be built with the customer group of identical general character, analyze user's owning user group, adopt collaborative filtering blending algorithm to produce and recommend list of songs, push and recommend list of songs to user.In whole process, carry out customer group accurately for huge user to divide, by excavating the degree of depth of user data, realize having an X-rayed the various dimensions of user interest, improve music recommend precision, in addition, build the customer group with identical general character, combine mixing proposed algorithm again, further improve accuracy and the reliability of music recommend.
Wherein in an embodiment, according to user personality, Similarity Measure is carried out to similar users set, builds user network and specifically comprise step:
Step one: analyze the frequency that user listens to often kind of song classification.
Types of songs is listened to user each in similar users set and carries out statistical study, analyze the frequency that user listens to often kind of song classification.
Step 2: reject the song classification lower than predeterminated frequency threshold value, and distribute song category IDs to remaining song classification.
Predeterminated frequency threshold value presets according to the needs of actual conditions, song classification lower than predetermined threshold value is rejected, the quantity of classification can be reduced, reduce late time data statistical study amount, avoid later stage mass data to be pushed to user and cause user to be difficult to select.After having screened, distribute song category IDs to remaining song classification, song category IDs, as the identification marking of song, can know the classification knowing current song by song category IDs.1 be Japan and Korea S, 2 be rock and roll in fig. 2,3 for rural area, 4 is for Guangdong language.
Step 3: scanning user history log, by abstract for each user be a node, and be each peer distribution node ID.
As shown in Figure 2, each user is abstract is a node, and each node is assigned node ID (node 1-18).
These two nodes are connected to straight line by step 4: when two nodes carried out operation to same song classification, wherein, carried out operation comprise collection or down operation to song classification.
When two nodes carried out operation to same song classification, namely showed to have two the different songs of user to identical category to operate, the operation of indication here included but not limited to collection, downloaded or purchase etc.Same song classification is operated, namely shows that different user is interested in identical category song, these two nodes are connected to straight line.
Step 5: repeat step 4, until scanned all users, has built user network.
Wherein in an example, described analysis user owning user group, adopts collaborative filtering blending algorithm to produce and recommends list of songs specifically to comprise step:
Construct the user-music-song classification three-dimensional matrice of each customer group;
Launch the three-dimensional matrice of user's owning user group, use user-song classification matrix to expand user-music matrix;
Launch the three-dimensional matrice of user's owning user group, use music-song classification matrix to expand user-music matrix;
Use the collaborative filtering blending algorithm based on user and music to produce to customer groups all belonging to user and recommend list of songs.
The process of " use the collaborative filtering blending algorithm based on user and music to produce to customer groups all belonging to user and recommend list of songs " will be described below in detail:
Matrix after expanding user-music matrix according to user-song classification matrix calculates the N number of user characteristics N of the highest Top of user characteristics similarity
u, pass through formula
calculate the contribution based on User Part;
Matrix after expanding user-music matrix according to music-song classification matrix, uses the Top-N algorithm based on song classification to obtain recommending song N
i, pass through formula
calculate the contribution based on musical portions;
Pass through formula
Calculate the fusion scoring based on user characteristics and song classification;
Choose the highest P song of scoring as last recommendation results.
As shown in Figure 3, one has an X-rayed music recommend system based on group, comprising:
Acquisition module 100, for obtaining user's history log, described user's history log record comprises user and listens to geographical location information residing for song title, user, user's use habit information, customer consumption ability information, individual subscriber preference information and age of user information;
Hobby category analysis module 200, for according to described user's history log, analyzes user in user personality and preset musical storehouse and listens to the hobby classification of song;
Customer group builds module 300, for listening to the hobby classification of song according to user in described user personality and preset musical storehouse, dividing user's set, being built with the customer group of identical general character;
List Generating Module 400, for analyzing user's owning user group, adopting collaborative filtering blending algorithm to produce and recommending list of songs;
Pushing module 500, for pushing described recommendation list of songs.
The present invention is based on group and have an X-rayed music recommend system, acquisition module 100 obtains user's history log, described user's history log is according to user's history log, hobby category analysis module 200 analyzes the hobby classification that user personality and user listen to song, customer group builds module 300 according to user personality and hobby classification, divide user's set, be built with the customer group of identical general character, List Generating Module 400 analyzes user's owning user group, adopt collaborative filtering blending algorithm to produce and recommend list of songs, pushing module 500 pushes recommends list of songs to user.In whole process, carry out customer group accurately for huge user to divide, by excavating the degree of depth of user data, realize having an X-rayed the various dimensions of user interest, improve music recommend precision, in addition, build the customer group with identical general character, combine mixing proposed algorithm again, further improve accuracy and the reliability of music recommend.
Wherein in an embodiment, described hobby category analysis module 200 specifically comprises:
Taxon, for classifying to the song in preset musical storehouse, and the title of song and quantity under recording every kind;
Analytic unit, for according to user's history log, analyzes user personality and obtains user and listen to song, analyzes user and listens to song generic;
Computing unit, listening to for calculating user the ratio that single classification number of songs and user listened to all number of songs, calculating the ratio of all number of songs in single classification number of songs and preset musical storehouse;
Hobby analytic unit, listened to the ratio of all number of songs in the ratio of all number of songs and single classification number of songs and preset musical storehouse for listening to single classification number of songs and user according to user, and analyzed user in preset musical storehouse and listen to the hobby classification of song.
Wherein in an embodiment, described customer group builds module 300 and specifically comprises:
A division unit, for according to user personality, once divides user's set, obtains preliminary user set;
Matching unit, for listening to the hobby classification of song according to user in preset musical storehouse, carries out fuzzy matching to described preliminary user set, obtains similar users set;
Similarity calculated, for according to user personality, carries out Similarity Measure to similar users set, builds user network;
Customer group construction unit, for adopting user network described in the community discovery Algorithm Analysis in complex network, is built with the customer group of identical general character.
Wherein in an embodiment, described similarity calculated specifically comprises:
Frequency analysis unit, listens to the frequency of often kind of song classification for analyzing user;
Screening unit, for rejecting the song classification lower than predeterminated frequency threshold value, and distributes song category IDs to remaining song classification;
ID allocation units, for scanning user's history log, by abstract for each user be a node, and be each peer distribution node ID;
These two nodes, for when two nodes carried out operation to same song classification, were connected to straight line by processing unit, wherein, carried out operation comprise collection or down operation to song classification;
Cycling element, for controlling, the repetition of described processing unit is described is connected to the operation of straight line when two nodes carried out operation to same song classification by these two nodes, until scanned all users, builds user network.
Wherein in an embodiment, described List Generating Module 400 specifically comprises:
Three-dimensional matrice tectonic element, for constructing the user-music-song classification three-dimensional matrice of each customer group;
First expanding unit, for launching the three-dimensional matrice of user's owning user group, uses user-song classification matrix to expand user-music matrix;
Second expanding unit, for launching the three-dimensional matrice of user's owning user group, uses music-song classification matrix to expand user-music matrix;
Recommendation list unit, uses the collaborative filtering blending algorithm based on user and music to produce recommendation list of songs for customer groups all belonging to user.
The above embodiment only have expressed several embodiment of the present invention, and it describes comparatively concrete and detailed, but can not therefore be construed as limiting the scope of the patent.It should be pointed out that for the person of ordinary skill of the art, without departing from the inventive concept of the premise, can also make some distortion and improvement, these all belong to protection scope of the present invention.Therefore, the protection domain of patent of the present invention should be as the criterion with claims.
Claims (10)
1. have an X-rayed a music recommend method based on group, it is characterized in that, comprise step:
Obtain user's history log, described user's history log record comprises user and listens to geographical location information residing for song title, user, user's use habit information, customer consumption ability information, individual subscriber preference information and age of user information;
According to described user's history log, analyze user in user personality and preset musical storehouse and listen to the hobby classification of song;
Listen to the hobby classification of song according to user in described user personality and preset musical storehouse, divide user's set, be built with the customer group of identical general character;
Analyze user's owning user group, adopt collaborative filtering blending algorithm to produce and recommend list of songs;
Push described recommendation list of songs.
2. according to claim 1ly have an X-rayed music recommend method based on group, it is characterized in that, described according to described user's history log, analyze the hobby classification that user in user personality and preset musical storehouse listens to song and specifically comprise step:
Song in preset musical storehouse is classified, and the title of song and quantity under recording every kind;
According to user's history log, analyze user personality and obtain user and listen to song, analyze user and listen to song generic;
Calculate user and listen to the ratio that single classification number of songs and user listened to all number of songs, calculate the ratio of all number of songs in single classification number of songs and preset musical storehouse;
Listen to according to user the ratio that single classification number of songs and user listened to all number of songs in the ratio of all number of songs and single classification number of songs and preset musical storehouse, analyze user in preset musical storehouse and listen to the hobby classification of song.
3. according to claim 1 and 2ly have an X-rayed music recommend method based on group, it is characterized in that, describedly listen to the hobby classification of song according to user in described user personality and preset musical storehouse, divide user's set, the customer group being built with identical general character specifically comprises step:
According to user personality, user's set is once divided, obtains preliminary user set;
Listen to the hobby classification of song according to user in preset musical storehouse, fuzzy matching is carried out to described preliminary user set, obtain similar users set;
According to user personality, Similarity Measure is carried out to similar users set, build user network;
Adopt user network described in the community discovery Algorithm Analysis in complex network, be built with the customer group of identical general character.
4. according to claim 3ly have an X-rayed music recommend method based on group, it is characterized in that, described according to user personality, Similarity Measure is carried out to similar users set, builds user network and specifically comprise step:
Analyze the frequency that user listens to often kind of song classification;
Reject the song classification lower than predeterminated frequency threshold value, and distribute song category IDs to remaining song classification;
Scanning user history log, by abstract for each user be a node, and be each peer distribution node ID;
When two nodes carried out operation to same song classification, these two nodes were connected to straight line, wherein, operation were carried out to song classification and comprises collection or down operation;
Described in repeating when two nodes carried out operation to same song classification, these two nodes are connected to the operation of straight line, until scanned all users, build user network.
5. according to claim 1 and 2ly have an X-rayed music recommend method based on group, it is characterized in that, described analysis user owning user group, adopt collaborative filtering blending algorithm to produce and recommend list of songs specifically to comprise step:
Construct the user-music-song classification three-dimensional matrice of each customer group;
Launch the three-dimensional matrice of user's owning user group, use user-song classification matrix to expand user-music matrix;
Launch the three-dimensional matrice of user's owning user group, use music-song classification matrix to expand user-music matrix;
Use the collaborative filtering blending algorithm based on user and music to produce to customer groups all belonging to user and recommend list of songs.
6. have an X-rayed a music recommend system based on group, it is characterized in that, comprising:
Acquisition module, for obtaining user's history log, described user's history log record comprises user and listens to geographical location information residing for song title, user, user's use habit information, customer consumption ability information, individual subscriber preference information and age of user information;
Hobby category analysis module, for according to described user's history log, analyzes user in user personality and preset musical storehouse and listens to the hobby classification of song;
Customer group builds module, for listening to the hobby classification of song according to user in described user personality and preset musical storehouse, dividing user's set, being built with the customer group of identical general character;
List Generating Module, for analyzing user's owning user group, adopting collaborative filtering blending algorithm to produce and recommending list of songs;
Pushing module, for pushing described recommendation list of songs.
7. according to claim 6ly have an X-rayed music recommend system based on group, it is characterized in that, described hobby category analysis module specifically comprises:
Taxon, for classifying to the song in preset musical storehouse, and the title of song and quantity under recording every kind;
Analytic unit, for according to user's history log, analyzes user personality and obtains user and listen to song, analyzes user and listens to song generic;
Computing unit, listening to for calculating user the ratio that single classification number of songs and user listened to all number of songs, calculating the ratio of all number of songs in single classification number of songs and preset musical storehouse;
Hobby analytic unit, listened to the ratio of all number of songs in the ratio of all number of songs and single classification number of songs and preset musical storehouse for listening to single classification number of songs and user according to user, and analyzed user in preset musical storehouse and listen to the hobby classification of song.
8. according to claim 6 or 7, have an X-rayed music recommend system based on group, it is characterized in that, described customer group builds module and specifically comprises:
A division unit, for according to user personality, once divides user's set, obtains preliminary user set;
Matching unit, for listening to the hobby classification of song according to user in preset musical storehouse, carries out fuzzy matching to described preliminary user set, obtains similar users set;
Similarity calculated, for according to user personality, carries out Similarity Measure to similar users set, builds user network;
Customer group construction unit, for adopting user network described in the community discovery Algorithm Analysis in complex network, is built with the customer group of identical general character.
9. according to claim 8ly have an X-rayed music recommend system based on group, it is characterized in that, described similarity calculated specifically comprises:
Frequency analysis unit, listens to the frequency of often kind of song classification for analyzing user;
Screening unit, for rejecting the song classification lower than predeterminated frequency threshold value, and distributes song category IDs to remaining song classification;
ID allocation units, for scanning user's history log, by abstract for each user be a node, and be each peer distribution node ID;
These two nodes, for when two nodes carried out operation to same song classification, were connected to straight line by processing unit, wherein, carried out operation comprise collection or down operation to song classification;
Cycling element, for controlling, the repetition of described processing unit is described is connected to the operation of straight line when two nodes carried out operation to same song classification by these two nodes, until scanned all users, builds user network.
10. according to claim 6 or 7, have an X-rayed music recommend system based on group, it is characterized in that, described List Generating Module specifically comprises:
Three-dimensional matrice tectonic element, for constructing the user-music-song classification three-dimensional matrice of each customer group;
First expanding unit, for launching the three-dimensional matrice of user's owning user group, uses user-song classification matrix to expand user-music matrix;
Second expanding unit, for launching the three-dimensional matrice of user's owning user group, uses music-song classification matrix to expand user-music matrix;
Recommendation list unit, uses the collaborative filtering blending algorithm based on user and music to produce recommendation list of songs for customer groups all belonging to user.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510153358.2A CN104731954B (en) | 2015-04-01 | 2015-04-01 | Music is had an X-rayed based on group and recommends method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510153358.2A CN104731954B (en) | 2015-04-01 | 2015-04-01 | Music is had an X-rayed based on group and recommends method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104731954A true CN104731954A (en) | 2015-06-24 |
| CN104731954B CN104731954B (en) | 2018-01-26 |
Family
ID=53455841
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510153358.2A Active CN104731954B (en) | 2015-04-01 | 2015-04-01 | Music is had an X-rayed based on group and recommends method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104731954B (en) |
Cited By (34)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105005612A (en) * | 2015-07-09 | 2015-10-28 | 广东欧珀移动通信有限公司 | Music file acquisition method and mobile terminal |
| CN106339504A (en) * | 2016-09-23 | 2017-01-18 | 乐视控股（北京）有限公司 | Music pushing method and apparatus |
| CN106484777A (en) * | 2016-09-12 | 2017-03-08 | 腾讯科技（深圳）有限公司 | A kind of multimedia data processing method and device |
| 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 |
| CN106649686A (en) * | 2016-12-16 | 2017-05-10 | 天翼爱音乐文化科技有限公司 | User interest group dividing method and system based on multilayer potential features |
| CN106844504A (en) * | 2016-12-27 | 2017-06-13 | 广州酷狗计算机科技有限公司 | A kind of method and apparatus for sending the single mark of song |
| CN106909620A (en) * | 2017-01-17 | 2017-06-30 | 广东小天才科技有限公司 | One kind teaches feature music assorting method and system for early |
| CN106951527A (en) * | 2017-03-21 | 2017-07-14 | 北京邮电大学 | A kind of song recommendations method and device |
| WO2017124386A1 (en) * | 2016-01-21 | 2017-07-27 | 阮元 | Method for acquiring technical data during time period-based resource recommendation, and recommendation system |
| WO2017124393A1 (en) * | 2016-01-21 | 2017-07-27 | 阮元 | Information pushing method during user preference-based resource recommendation, and recommendation system |
| WO2017124387A1 (en) * | 2016-01-21 | 2017-07-27 | 阮元 | Information pushing method during time period-based automatic resource recommendation, and recommendation system |
| WO2017124380A1 (en) * | 2016-01-21 | 2017-07-27 | 阮元 | Method for collecting statistics about technology usage data during audio resource recommendation, and recommendation system |
| WO2017173572A1 (en) * | 2016-04-05 | 2017-10-12 | 张阳 | Music control method and system for disco club |
| CN107426338A (en) * | 2017-08-28 | 2017-12-01 | 江西博瑞彤芸科技有限公司 | A kind of information management method and system |
| WO2018018190A1 (en) * | 2016-07-24 | 2018-02-01 | 严映军 | Data acquisition method for region-based program recommendation and program playback system |
| WO2018018192A1 (en) * | 2016-07-24 | 2018-02-01 | 严映军 | Region-based program recommendation method and program playback system |
| WO2018018191A1 (en) * | 2016-07-24 | 2018-02-01 | 严映军 | Information pushing method during birthplace-program matching and program playback system |
| CN107908701A (en) * | 2017-11-06 | 2018-04-13 | 广东欧珀移动通信有限公司 | Method, apparatus, storage medium and the terminal device that music is recommended |
| CN108009181A (en) * | 2016-10-28 | 2018-05-08 | 北京酷我科技有限公司 | Recommend method and system in a kind of radio station |
| CN108021574A (en) * | 2016-11-02 | 2018-05-11 | 北京酷我科技有限公司 | A kind of searching method and device |
| CN108205551A (en) * | 2016-12-16 | 2018-06-26 | 北京酷我科技有限公司 | A kind of song recommendations method and song recommendations system |
| CN108959319A (en) * | 2017-05-25 | 2018-12-07 | 腾讯科技（深圳）有限公司 | Information-pushing method and device |
| CN109271550A (en) * | 2018-07-27 | 2019-01-25 | 华南理工大学 | A kind of music personalization classification recommended method based on deep learning |
| CN109559140A (en) * | 2017-09-26 | 2019-04-02 | 广州市动景计算机科技有限公司 | It expands the other method, apparatus of user interest degree group, calculate equipment and storage medium |
| CN110275980A (en) * | 2019-06-26 | 2019-09-24 | 徐州工业职业技术学院 | One kind having an X-rayed music recommended method based on group |
| 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 |
| CN110309346A (en) * | 2018-03-02 | 2019-10-08 | 上海博泰悦臻网络技术服务有限公司 | Song recommendations method, sound system and server based on indoor location positioning |
| CN111026957A (en) * | 2019-11-21 | 2020-04-17 | 北京网聘咨询有限公司 | Recommendation system and method based on multi-dimensional similarity |
| CN111488485A (en) * | 2020-04-16 | 2020-08-04 | 北京雷石天地电子技术有限公司 | Music recommendation method based on convolutional neural network, storage medium and electronic device |
| CN111723234A (en) * | 2020-06-15 | 2020-09-29 | 中国第一汽车股份有限公司 | Audio providing method, device, equipment and storage medium |
| CN112732971A (en) * | 2021-01-21 | 2021-04-30 | 广西师范大学 | Collaborative filtering music recommendation method based on labels |
| CN113286199A (en) * | 2020-02-20 | 2021-08-20 | 佛山市云米电器科技有限公司 | Program recommendation method, television and storage medium |
| CN115022363A (en) * | 2022-05-30 | 2022-09-06 | 深圳季连科技有限公司 | Information sharing method considering safety based on Internet of vehicles |
| CN115065845A (en) * | 2022-06-17 | 2022-09-16 | 联通沃音乐文化有限公司 | Video and audio playing management method and system for community fusion recommendation |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090077132A1 (en) * | 2005-09-28 | 2009-03-19 | Sony Corporation | Information Processing Device and Method, and Program |
| CN101431485A (en) * | 2008-12-31 | 2009-05-13 | 深圳市迅雷网络技术有限公司 | Method and system for automatically recommending internet information |
| 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 |
| CN102637178A (en) * | 2011-02-14 | 2012-08-15 | 北京瑞信在线系统技术有限公司 | Music recommending method, music recommending device and music recommending system |
| CN102654859A (en) * | 2011-03-01 | 2012-09-05 | 北京彩云在线技术开发有限公司 | Method and system for recommending songs |
| CN103559197A (en) * | 2013-09-23 | 2014-02-05 | 浙江大学 | Real-time music recommendation method based on context pre-filtering |
| CN104008184A (en) * | 2014-06-10 | 2014-08-27 | 百度在线网络技术（北京）有限公司 | Method and device for pushing information |
-
2015
- 2015-04-01 CN CN201510153358.2A patent/CN104731954B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090077132A1 (en) * | 2005-09-28 | 2009-03-19 | Sony Corporation | Information Processing Device and Method, and Program |
| CN101431485A (en) * | 2008-12-31 | 2009-05-13 | 深圳市迅雷网络技术有限公司 | Method and system for automatically recommending internet information |
| CN102637178A (en) * | 2011-02-14 | 2012-08-15 | 北京瑞信在线系统技术有限公司 | Music recommending method, music recommending device and music recommending system |
| CN102654859A (en) * | 2011-03-01 | 2012-09-05 | 北京彩云在线技术开发有限公司 | Method and system for recommending songs |
| 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 |
| CN103559197A (en) * | 2013-09-23 | 2014-02-05 | 浙江大学 | Real-time music recommendation method based on context pre-filtering |
| CN104008184A (en) * | 2014-06-10 | 2014-08-27 | 百度在线网络技术（北京）有限公司 | Method and device for pushing information |
Non-Patent Citations (1)
| Title |
|---|
| KAREN H. L. TSOSUTTER等: "Tag-aware Recommender Systems by Fusion of Collaborative Filtering Algorithms", 《PROCESSING OF THE 2008 ACM SYMPOSIUM ON APPLIED COMPUTING》 * |
Cited By (44)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105005612B (en) * | 2015-07-09 | 2018-11-20 | 广东欧珀移动通信有限公司 | A kind of acquisition methods and mobile terminal of music file |
| CN105005612A (en) * | 2015-07-09 | 2015-10-28 | 广东欧珀移动通信有限公司 | Music file acquisition method and mobile terminal |
| WO2017124380A1 (en) * | 2016-01-21 | 2017-07-27 | 阮元 | Method for collecting statistics about technology usage data during audio resource recommendation, and recommendation system |
| WO2017124387A1 (en) * | 2016-01-21 | 2017-07-27 | 阮元 | Information pushing method during time period-based automatic resource recommendation, and recommendation system |
| WO2017124393A1 (en) * | 2016-01-21 | 2017-07-27 | 阮元 | Information pushing method during user preference-based resource recommendation, and recommendation system |
| WO2017124386A1 (en) * | 2016-01-21 | 2017-07-27 | 阮元 | Method for acquiring technical data during time period-based resource recommendation, and recommendation system |
| WO2017173572A1 (en) * | 2016-04-05 | 2017-10-12 | 张阳 | Music control method and system for disco club |
| WO2018018191A1 (en) * | 2016-07-24 | 2018-02-01 | 严映军 | Information pushing method during birthplace-program matching and program playback system |
| WO2018018190A1 (en) * | 2016-07-24 | 2018-02-01 | 严映军 | Data acquisition method for region-based program recommendation and program playback system |
| WO2018018192A1 (en) * | 2016-07-24 | 2018-02-01 | 严映军 | Region-based program recommendation method and program playback system |
| CN106484777A (en) * | 2016-09-12 | 2017-03-08 | 腾讯科技（深圳）有限公司 | A kind of multimedia data processing method and device |
| CN106339504A (en) * | 2016-09-23 | 2017-01-18 | 乐视控股（北京）有限公司 | Music pushing method and apparatus |
| CN108009181B (en) * | 2016-10-28 | 2021-03-19 | 北京酷我科技有限公司 | Radio station recommendation method and system |
| CN108009181A (en) * | 2016-10-28 | 2018-05-08 | 北京酷我科技有限公司 | Recommend method and system in a kind of radio station |
| CN108021574A (en) * | 2016-11-02 | 2018-05-11 | 北京酷我科技有限公司 | A kind of searching method and device |
| 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 |
| CN108205551B (en) * | 2016-12-16 | 2020-09-29 | 北京酷我科技有限公司 | Song recommendation method and song recommendation system |
| CN106649686A (en) * | 2016-12-16 | 2017-05-10 | 天翼爱音乐文化科技有限公司 | User interest group dividing method and system based on multilayer potential features |
| CN108205551A (en) * | 2016-12-16 | 2018-06-26 | 北京酷我科技有限公司 | A kind of song recommendations method and song recommendations system |
| CN106844504B (en) * | 2016-12-27 | 2019-10-11 | 广州酷狗计算机科技有限公司 | A kind of method and apparatus for sending song and singly identifying |
| CN106844504A (en) * | 2016-12-27 | 2017-06-13 | 广州酷狗计算机科技有限公司 | A kind of method and apparatus for sending the single mark of song |
| CN106909620A (en) * | 2017-01-17 | 2017-06-30 | 广东小天才科技有限公司 | One kind teaches feature music assorting method and system for early |
| CN106909620B (en) * | 2017-01-17 | 2020-07-28 | 广东小天才科技有限公司 | Functional music classification method and system for early education |
| CN106951527B (en) * | 2017-03-21 | 2020-01-17 | 北京邮电大学 | Song recommendation method and device |
| CN106951527A (en) * | 2017-03-21 | 2017-07-14 | 北京邮电大学 | A kind of song recommendations method and device |
| CN108959319A (en) * | 2017-05-25 | 2018-12-07 | 腾讯科技（深圳）有限公司 | Information-pushing method and device |
| CN107426338A (en) * | 2017-08-28 | 2017-12-01 | 江西博瑞彤芸科技有限公司 | A kind of information management method and system |
| CN109559140A (en) * | 2017-09-26 | 2019-04-02 | 广州市动景计算机科技有限公司 | It expands the other method, apparatus of user interest degree group, calculate equipment and storage medium |
| CN107908701A (en) * | 2017-11-06 | 2018-04-13 | 广东欧珀移动通信有限公司 | Method, apparatus, storage medium and the terminal device that music is recommended |
| CN110309346A (en) * | 2018-03-02 | 2019-10-08 | 上海博泰悦臻网络技术服务有限公司 | Song recommendations method, sound system and server based on indoor location positioning |
| CN109271550A (en) * | 2018-07-27 | 2019-01-25 | 华南理工大学 | A kind of music personalization classification recommended method based on deep learning |
| CN109271550B (en) * | 2018-07-27 | 2022-05-24 | 华南理工大学 | Music personalized recommendation method based on deep learning |
| 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 |
| CN110275980A (en) * | 2019-06-26 | 2019-09-24 | 徐州工业职业技术学院 | One kind having an X-rayed music recommended method based on group |
| CN111026957A (en) * | 2019-11-21 | 2020-04-17 | 北京网聘咨询有限公司 | Recommendation system and method based on multi-dimensional similarity |
| CN111026957B (en) * | 2019-11-21 | 2023-05-05 | 北京网聘咨询有限公司 | Recommendation system and method based on multidimensional similarity |
| CN113286199A (en) * | 2020-02-20 | 2021-08-20 | 佛山市云米电器科技有限公司 | Program recommendation method, television and storage medium |
| CN111488485A (en) * | 2020-04-16 | 2020-08-04 | 北京雷石天地电子技术有限公司 | Music recommendation method based on convolutional neural network, storage medium and electronic device |
| CN111488485B (en) * | 2020-04-16 | 2023-11-17 | 北京雷石天地电子技术有限公司 | Music recommendation method based on convolutional neural network, storage medium and electronic device |
| CN111723234A (en) * | 2020-06-15 | 2020-09-29 | 中国第一汽车股份有限公司 | Audio providing method, device, equipment and storage medium |
| CN112732971A (en) * | 2021-01-21 | 2021-04-30 | 广西师范大学 | Collaborative filtering music recommendation method based on labels |
| CN115022363A (en) * | 2022-05-30 | 2022-09-06 | 深圳季连科技有限公司 | Information sharing method considering safety based on Internet of vehicles |
| CN115022363B (en) * | 2022-05-30 | 2024-04-16 | 深圳季连科技有限公司 | Information sharing method based on internet of vehicles and considering safety |
| CN115065845A (en) * | 2022-06-17 | 2022-09-16 | 联通沃音乐文化有限公司 | Video and audio playing management method and system for community fusion recommendation |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104731954B (en) | 2018-01-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104731954A (en) | Music recommendation method and system based on group perspective | |
| CN110245981B (en) | Crowd type identification method based on mobile phone signaling data | |
| CN107577688B (en) | Original article influence analysis system based on media information acquisition | |
| CN109271477B (en) | Method and system for constructing classified corpus by means of Internet | |
| CN102024058B (en) | Music recommendation method and system | |
| CN101119326B (en) | Method and device for managing instant communication conversation record | |
| US20120174006A1 (en) | System, method, apparatus and computer program for generating and modeling a scene | |
| CN103544188B (en) | The user preference method for pushing of mobile Internet content and device | |
| CN106339502A (en) | Modeling recommendation method based on user behavior data fragmentation cluster | |
| CN109271550B (en) | Music personalized recommendation method based on deep learning | |
| CN105426514A (en) | Personalized mobile APP recommendation method | |
| CN101984437A (en) | Music resource individual recommendation method and system thereof | |
| CN102637178A (en) | Music recommending method, music recommending device and music recommending system | |
| US10387805B2 (en) | System and method for ranking news feeds | |
| CN101820592A (en) | Method and device for mobile search | |
| CN104573130A (en) | Entity resolution method based on group calculation and entity resolution device based on group calculation | |
| CN104573114A (en) | Music classification method and device | |
| CN111523055A (en) | Collaborative recommendation method and system based on agricultural product characteristic attribute comment tendency | |
| CN103455538A (en) | Information processing apparatus, information processing method, and program | |
| CN111984873A (en) | Service recommendation system and method | |
| CN102156747A (en) | Method and device for forecasting collaborative filtering mark by introduction of social tag | |
| CN114077705A (en) | Method and system for portraying media account on social platform | |
| CN101788987A (en) | Automatic judging method of network resource types | |
| Gulmatico et al. | SpotiPred: A machine learning approach prediction of Spotify music popularity by audio features | |
| CN116881430A (en) | Industrial chain identification method and device, electronic equipment and readable storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| CB03 | Change of inventor or designer information |
Inventor after: Zhu Yingbo Inventor after: Zeng Rong Inventor after: Wu Weifen Inventor after: Shi Yipeng Inventor after: Pan Zhifeng Inventor before: Zeng Rong Inventor before: Wu Weifen |
|
| COR | Change of bibliographic data | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |