CN109992694A - A kind of music intelligent recommendation method and system - Google Patents
A kind of music intelligent recommendation method and system Download PDFInfo
- Publication number
- CN109992694A CN109992694A CN201910306677.0A CN201910306677A CN109992694A CN 109992694 A CN109992694 A CN 109992694A CN 201910306677 A CN201910306677 A CN 201910306677A CN 109992694 A CN109992694 A CN 109992694A
- Authority
- CN
- China
- Prior art keywords
- song
- similarity
- candidate
- lyrics
- recommended
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Abstract
The invention discloses a kind of music intelligent recommendation method and system, comprising: obtains the song that user likes and is used as with reference to song;The melody characteristics for referring to song are extracted, the type for referring to song is extracted, extract the lyrics for referring to song;The melody characteristics of reference song are matched with the melody characteristics of song in library, obtain melodic similarity, melodic similarity is selected as candidate song when being higher than default similarity from library;Extract the type of candidate song;Select song identical with reference types of songs as song to be recommended from candidate song;The lyrics of reference song and the lyrics of song to be recommended are compared, song to be recommended is ranked up and is presented to the user according to lyrics similarity.A kind of music intelligent recommendation method and system provided by the invention, can be found according to the favorite song of user institute by the favorite song of user institute and can recommend user, improve and recommend accuracy.
Description
Technical field
The present invention relates to music recommended technology fields, and in particular to a kind of music intelligent recommendation method and system.
Background technique
With the development of information technology and network technology, music libraries scale increases by geometric progression, the type of music song
Also increase therewith, more and more users are by internet or mobile Internet is listened to online or down-load music.User is requesting a song
When, it is desirable to provide relevant information of music, such as title, the author of music etc., then server return meet search condition
Music list is selected for user.In addition, user may like certain a kind of music, it is therefore desirable to which music is actively recommended use
Family.Traditional music recommended method is pushed away only by music-related information (such as album name, author, type etc.) Lai Jinhang
It recommends, such as user listens to music A, music A and music B and belongs to same album, then it is assumed that user can also like music B, therefore
Music B is recommended into user.
However, traditional this music recommended method is only with reference to music-related information due to being recommended, but in reality
In, the habit difference of some users is very big, and often identical two music of relevant information can't be liked by user, because
The accuracy that this traditional this music recommended method is recommended is not high.
Summary of the invention
The present invention provides a kind of music intelligent recommendation method and system, and can be found according to the favorite song of user institute can
By the favorite song of user institute and user is recommended, improves and recommends accuracy.
In a first aspect, the present invention provides a kind of music intelligent recommendation method, which comprises
The song that user likes is obtained to be used as with reference to song;
The melody characteristics for referring to song are extracted, the type for referring to song is extracted, extract the lyrics for referring to song;
The melody characteristics of reference song are matched with the melody characteristics of song in library, obtain melodic similarity,
Melodic similarity is selected as candidate song when being higher than default similarity from library;
Extract the type of candidate song;
Select song identical with reference types of songs as song to be recommended from candidate song;
The lyrics of reference song and the lyrics of song to be recommended are compared, song to be recommended is carried out according to lyrics similarity
It sorts and is presented to the user.
Preferably, the song that the acquisition user likes, which is used as with reference to song, includes:
The song of user's selected broadcasting when opening music software is selected, and broadcasting frequency is more than n times;Or selection
The song recommendations of user's broadcasting time front three when wearing earphone are selected to user, refer to song to determine.
Preferably, the melody characteristics include: that the melody characteristics that sequence of notes indicates or the melody that pitch contour indicates are special
Sign;
Preferably, the melody characteristics by reference song are matched with the melody characteristics of song in library, are obtained
Melodic similarity includes that when being higher than default similarity, be selected as candidate song from library includes: melodic similarity
For the melody characteristics that sequence of notes indicates, calculated in the melody characteristics and library using sequences match algorithm
The similarity of the melody characteristics of song;
For the melody characteristics that pitch contour indicates, the melody characteristics and song are calculated using dynamic time warping algorithm
The melody characteristics similarity of song in library;
Calculate the similarity of the sequence of notes of song to be matched in the sequence of notes with reference to song and library;
Calculate the similarity of the pitch contour of song to be matched in the pitch contour and library with reference to song;
The similarity of the pitch contour is merged with the similarity of sequence of notes, and using fusion results as melody
Similarity;
Select melodic similarity be greater than given threshold song as candidate song, or according to melodic similarity by greatly to
The song of small sequential selection setting number is as candidate song.
Preferably, described to select song identical with reference types of songs as song packet to be recommended from candidate song
It includes:
It is described with reference to types of songs it is identical with the candidate song type or it is described refer to types of songs with it is described
Candidate song type is not exactly the same but has a kind of and above type identical.
Preferably, the lyrics by reference song and the lyrics of song to be recommended compare, and treat according to lyrics similarity
Recommendation song, which is ranked up and is presented to the user, includes:
The lyrics of the lyrics of reference song and song to be recommended are divided into logical sequence of terms using segmenter, it will
Traversal is carried out with the sequence of terms of song to be recommended with reference to the sequence of terms of song to compare, and obtains identical word quantity and different words
Language quantity;
According to the ratio of identical the word quantity and different terms quantity, the similarity of the lyrics is obtained;
Song to be recommended is ranked up in the way of from high to low by similarity according to lyrics similarity.
Second aspect, the present invention also provides a kind of music intelligent recommendation system, the system comprises: obtain module, ginseng
Examine song extraction module, selecting module, candidate song extraction module, song module to be recommended, sorting module.
Module is obtained, is used as obtaining the song that user likes with reference to song;
With reference to song extraction module, for extracting the melody characteristics, type and the lyrics that refer to song;
Selecting module, for selecting candidate song from library according to the melody characteristics similarity;The selection mould
Block includes: melody characteristics matching module, and the melody characteristics for that will refer to song in the melody characteristics and library of song carry out
Matching, obtains melodic similarity;Candidate block is selected as selecting from library when melodic similarity is when being higher than default similarity
Candidate song;
Candidate song extraction module, for extracting the type of candidate song, the type of candidate song is two kinds or more;
Song module to be recommended, for selecting song identical with reference types of songs as to be recommended from candidate song
Song;
Sorting module, for the lyrics of song and the lyrics comparison of song to be recommended will to be referred to, according to lyrics similarity pair
Song to be recommended is ranked up and is presented to the user.
Preferably, the melody characteristics matching module includes:
First computing unit matches the note of song with band in library for calculating the sequence of notes with reference to song
The similarity of each sub- sequence of notes in sequence;
Second computing unit matches the fundamental frequency of song with band in library for calculating the pitch contour with reference to song
The similarity of envelope;
Integrated unit, for merging the similarity of the pitch contour with the similarity of sequence of notes;
Output unit, the fusion results for obtaining the integrated unit are exported as melodic similarity.
The candidate block specially selects melodic similarity greater than the song of given threshold as candidate song, or
According to the song of the descending sequential selection setting number of melodic similarity as candidate song.
A kind of music intelligent recommendation method and system provided by the invention select user is selected when opening music software to broadcast
The song put, and the higher song of broadcasting frequency is used as and refers to song, and according to daily habits, which is most to benefit from the recent period
The favorite song in family.By the melody of music, type and the lyrics scan for simultaneously, improve the song searched by user institute
The accuracy liked.
Detailed description of the invention
Fig. 1 is the flow diagram of music intelligent recommendation method provided by the invention.
Fig. 2 is the structural schematic diagram of music intelligent recommendation system provided by the invention.
Fig. 3 is the structural schematic diagram of the melody characteristics matching module of music intelligent recommendation system provided by the invention
Specific embodiment
The present invention is described in further detail with reference to the accompanying drawing.
Embodiment one
Fig. 1 shows the flow diagram of the music intelligent recommendation method of the offer of the embodiment of the present invention one.The method packet
It includes
Step S1 obtains the song that user likes and is used as with reference to song;
Step S2 extracts the melody characteristics for referring to song, extracts the type for referring to song, extracts the lyrics for referring to song;
The melody characteristics of reference song are matched with the melody characteristics of song in library, obtain melody by step S3
Similarity, melodic similarity are selected as candidate song when being higher than default similarity from library;
Step S4 extracts the type of candidate song;
Step S5 selects song identical with reference types of songs as song to be recommended from candidate song;
Step S6 compares the lyrics of reference song and the lyrics of song to be recommended, according to lyrics similarity to be recommended
Song is ranked up and is presented to the user.
The embodiment of the present invention one the specific technical proposal is:
Step S1 obtains the song that user likes and is used as with reference to song;
The song of user's selected broadcasting when opening music software is selected, and broadcasting frequency is at 3 times or more;Or selection
The song recommendations of user's broadcasting time front three when wearing earphone are selected to user, refer to song to determine.
Step S2 extracts the melody characteristics for referring to song, extracts the type for referring to song, extracts the lyrics for referring to song;
Melody characteristics include: the melody characteristics that the melody characteristics that sequence of notes indicates or pitch contour indicate;
The melody characteristics of reference song are matched with the melody characteristics of song in library, obtain melody by step S3
Similarity, melodic similarity are selected as candidate song when being higher than default similarity from library;
In embodiments of the present invention, the melody characteristics can have sequence of notes or pitch contour to characterize.
For the melody characteristics that sequence of notes indicates, sequences match algorithm can use, such as ED (Edit Distance,
Editing distance) algorithm, LCS (Longest Common Subsequence, longest common subsequence) algorithm etc., it calculates with reference to song
The similarity of the melody characteristics of song in bent melody characteristics and library.
For the melody characteristics that pitch contour indicates, DTW (Dynamic Time Warping, dynamic time can use
It is regular) algorithm calculate with reference to song melody characteristics and library in song melody characteristics similarity.
Calculate in sequence of notes and library with reference to song each sub- sequence of notes in the sequence of notes of song to be matched
Similarity;
Calculate the similarity of the pitch contour of song to be matched in pitch contour and library with reference to song;
The similarity of the pitch contour is merged with the similarity of the corresponding sub- sequence of notes of maximum, and will fusion
As a result it is used as melodic similarity;
Specific amalgamation mode can carry out as follows:
Sim=α S1+β·S2
Wherein, Sim indicates melody characteristics similarity, S1Indicate sequence of notes similarity, S2Indicate pitch contour similarity, α
It is fusion coefficients with β.
The melody characteristics of reference song are matched with the melody characteristics of song, and according to matching result from library
The high song of some similarities is selected as candidate song, for example, can choose the song that melodic similarity is greater than given threshold
As candidate song;Or the song conduct according to descending sequential selection setting number (such as 20) of melodic similarity
Candidate song.
Step S4 extracts the type of candidate song, in the present embodiment, the type of candidate song include it is existing it is cruel I, it is cruel
The classification standard of the music players such as dog, qq, such as " heat song ", " popular song ", " first signature song ", " classics are missed old times or old friends ", " jazz ",
" rock and roll " etc., song have index and define multiple labels;
The type of candidate song is two kinds or more, is also " love after 80s is listened " if candidate song is " popular song ".
Step S5 selects song identical with reference types of songs as song to be recommended from candidate song;
It is identical with the candidate song type or described with reference to types of songs and the candidate with reference to types of songs
Types of songs is not exactly the same but has a kind of and above type identical.
Step S6 compares the lyrics of reference song and the lyrics of song to be recommended, according to lyrics similarity to be recommended
Song is ranked up and is presented to the user.
The lyrics of reference song are divided into logical sequence of terms An={ word a1, word a2, word using segmenter
Language a3 ... ..., word an };
Using segmenter by the lyrics of song to be recommended be divided into logical sequence of terms Bn=word b1, word b2,
Word b3 ... ..., word bn };
The sequence of terms Bn of the sequence of terms An of reference song and song to be recommended are carried out traversal to compare, obtain same words
Language quantity N and different terms quantity M;
According to the ratio (N/M × 100%) of the identical word quantity and different terms quantity, the similar of the lyrics is obtained
Degree;
Song to be recommended is ranked up in the way of from high to low by similarity according to lyrics similarity.
Based on the above content, technical effect that the embodiment of the present invention one may be implemented are as follows: choose user and open music software
When selected broadcasting song, and the higher song of broadcasting frequency is used as and refers to song, quickly finds user with this and is liked
The song of love, and be that user quickly searches from library and can use according to melody, type and the lyrics with reference to song
The favorite song in family improves the accuracy that music is recommended.
Embodiment two
Accordingly to the embodiment of the present invention one, Fig. 2 shows music intelligent recommendation systems provided in an embodiment of the present invention
Structural schematic diagram.The system comprises: module 101 is obtained, with reference to song extraction module 102, selecting module 103, candidate song
Extraction module 104, song module 105 to be recommended, sorting module 106.
The acquisition module 101 is used as obtaining the song that user likes with reference to song.User is selected to open music soft
The song of selected broadcasting when part, and broadcasting frequency is at 3 times or more;Or selection user played when wearing earphone it is secondary
The song recommendations of number front three are selected to user, refer to song to determine.
It is described to refer to song extraction module 102, for extracting the melody characteristics, type and the lyrics that refer to song;
The selecting module 103, for selecting candidate song from library according to the melody characteristics similarity;It is described
Selecting module includes:
Melody characteristics matching module, the melody characteristics for that will refer to song in the melody characteristics and library of song carry out
Matching, obtains melodic similarity;The module includes the first computing unit, for calculating the sequence of notes and song with reference to song
The similarity of each sub- sequence of notes, utilizes sequences match algorithm, such as ED (Edit in sequence of notes with matching song in song library
Distance, editing distance) algorithm, LCS (Longest Common Subsequence, longest common subsequence) algorithm etc.,
Calculate the similarity of the melody characteristics of the song in the melody characteristics and library with reference to song;Second computing unit, based on
The similarity that the pitch contour with reference to song matches the pitch contour of song with band in library is calculated, DTW is utilized
(Dynamic Time Warping, dynamic time warping) algorithm calculates the song in melody characteristics and library with reference to song
Melody characteristics similarity;Integrated unit, for carrying out the similarity of the similarity of the pitch contour and sequence of notes
Fusion, fusion formula: Sim=α S1+ β S2, wherein Sim indicates melody characteristics similarity, and S1 indicates that sequence of notes is similar
Degree, S2 indicate pitch contour similarity, and α and β are fusion coefficients;Output unit, the fusion for obtaining the integrated unit
As a result it is exported as melodic similarity.
Candidate block is selected as selecting candidate song from library when melodic similarity is when being higher than default similarity.It will ginseng
The melody characteristics for examining song are matched with the melody characteristics of song, and are selected from library according to matching result some similar
High song is spent as candidate song, for example, can choose song of the melodic similarity greater than given threshold as candidate song;
Or the song of number (such as 20) is set as candidate song according to the descending sequential selection of melodic similarity.
Candidate song extraction module 104, for extracting the type of candidate song;In the present embodiment, the class of candidate song
Type includes existing cruel I, the classification standards of the music players such as KuGoo, qq, such as " heat song ", " popular song ", " first signature song ", " warp
Allusion quotation is missed old times or old friends ", " jazz ", " rock and roll " etc., song has index and defines multiple labels;The type of candidate song be two kinds and with
On, it is also " love after 80s is listened " if candidate song is " popular song ".
Song module 105 to be recommended, for select from candidate song with refer to the identical song of types of songs as to
Recommend song;It is identical with the candidate song type or described with reference to types of songs and the time with reference to types of songs
It selects types of songs not exactly the same but has a kind of and above type identical.
Sorting module 106, for the lyrics of song and the lyrics comparison of song to be recommended will to be referred to, according to lyrics similarity
Song to be recommended is ranked up and is presented to the user.The lyrics of reference song are divided into logical word using segmenter
The lyrics of song to be recommended are divided into logical sequence of terms Bn by sequence An;By the sequence of terms An of reference song with to
Recommend the sequence of terms Bn of song to carry out traversal comparison, obtains identical word quantity N and different terms quantity M;According to the phase
With the ratio (N/M × 100%) of word quantity and different terms quantity, the similarity of the lyrics is obtained;According to lyrics similarity pair
Song to be recommended is ranked up in the way of from high to low by similarity.
Based on the above content, what the embodiment of the present invention two can achieve has the technical effect that melody characteristics matching module uses
Different algorithm calculates the similarity of sequence of notes and the similarity of pitch contour is simultaneously merged as melodic similarity, improve according to
According to the accuracy that melodic similarity is searched for, and the song of same type is filtered out using song module to be recommended, is passing through sequence
Module is ranked up song to be recommended, in order to which user quickly hears the song of closest same type.
Finally, it should be noted that the above embodiments are merely illustrative of the technical solutions of the present invention, rather than its limitations;Although
Present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that;It still may be used
To modify to technical solution documented by previous embodiment, or some or all of the technical features are equal
Replacement;And these are modified or replaceed, the model for technical solution of the embodiment of the present invention that it does not separate the essence of the corresponding technical solution
It encloses, should all cover within the scope of the claims and the description of the invention.
Claims (8)
1. a kind of music intelligent recommendation method characterized by comprising
The song that user likes is obtained to be used as with reference to song;
The melody characteristics for referring to song are extracted, the type for referring to song is extracted, extract the lyrics for referring to song;
The melody characteristics of reference song are matched with the melody characteristics of song in library, obtain melodic similarity, melody
Similarity is selected as candidate song when being higher than default similarity from library;
Extract the type of candidate song;
Select song identical with reference types of songs as song to be recommended from candidate song;
The lyrics of reference song and the lyrics of song to be recommended are compared, song to be recommended is ranked up according to lyrics similarity
And it is presented to the user.
2. a kind of music intelligent recommendation method according to claim 1, which is characterized in that the song for obtaining user and liking
Qu Zuowei includes: with reference to song
The song of user's selected broadcasting when opening music software is selected, and broadcasting frequency is more than n times;Or selection user
When wearing earphone, the song recommendations of broadcasting time front three are selected to user, refer to song to determine.
3. a kind of music intelligent recommendation method according to claim 1, which is characterized in that the melody characteristics include: sound
Accord with the melody characteristics that sequence indicates or the melody characteristics that pitch contour indicates.
4. a kind of music intelligent recommendation method according to claim 1, which is characterized in that the melody by reference song
Feature is matched with the melody characteristics of song in library, is obtained melodic similarity and is included, and melodic similarity is higher than default
When similarity, being selected as candidate song from library includes:
For the melody characteristics that sequence of notes indicates, song in the melody characteristics and library is calculated using sequences match algorithm
Melody characteristics similarity;
For the melody characteristics that pitch contour indicates, calculated in the melody characteristics and library using dynamic time warping algorithm
The melody characteristics similarity of song;
Calculate the similarity of the sequence of notes of song to be matched in the sequence of notes with reference to song and library;
Calculate the similarity of the pitch contour of song to be matched in the pitch contour and library with reference to song;
The similarity of the pitch contour is merged with the similarity of sequence of notes, and similar using fusion results as melody
Degree;
The song for selecting melodic similarity to be greater than given threshold is descending as candidate song, or according to melodic similarity
Sequential selection sets the song of number as candidate song.
5. a kind of music intelligent recommendation method according to claim 1, which is characterized in that described to be selected from candidate song
Song identical with reference types of songs includes: as song to be recommended
It is described to refer to types of songs identical with the candidate song type or the reference types of songs and the candidate
Types of songs is not exactly the same but has a kind of and above type identical.
6. a kind of music intelligent recommendation method according to claim 1, which is characterized in that the lyrics by reference song
It is compared with the lyrics of song to be recommended, foundation lyrics similarity, which is ranked up and is presented to the user to song to be recommended, includes:
The lyrics of the lyrics of reference song and song to be recommended are divided into logical sequence of terms using segmenter, will be referred to
The sequence of terms of the sequence of terms of song and song to be recommended carries out traversal and compares, and obtains identical word quantity and different terms number
Amount;
According to the ratio of identical the word quantity and different terms quantity, the similarity of the lyrics is obtained;
Song to be recommended is ranked up in the way of from high to low by similarity according to lyrics similarity.
7. a kind of music intelligent recommendation system, which is characterized in that the system comprises: module is obtained, with reference to song extraction module,
Selecting module, candidate song extraction module, song module to be recommended, sorting module;
Module is obtained, is used as obtaining the song that user likes with reference to song;
With reference to song extraction module, for extracting the melody characteristics, type and the lyrics that refer to song;
Selecting module, for selecting candidate song from library according to the melody characteristics similarity;The selecting module packet
It includes: melody characteristics matching module, for the melody characteristics for referring to song to be matched with the melody characteristics of song in library,
Obtain melodic similarity;Candidate block is selected as selecting candidate song from library when melodic similarity is when being higher than default similarity
It is bent;
Candidate song extraction module, for extracting the type of candidate song;
Song module to be recommended, for selecting song identical with reference types of songs as song to be recommended from candidate song
It is bent;
Sorting module is treated according to lyrics similarity and is pushed away for that will refer to the lyrics of song and the lyrics comparison of song to be recommended
Song is recommended to be ranked up and be presented to the user.
8. a kind of music intelligent recommendation system according to claim 7, which is characterized in that the melody characteristics matching module
Include:
First computing unit matches the sequence of notes of song with band in library for calculating the sequence of notes with reference to song
In each sub- sequence of notes similarity;
Second computing unit, for calculating the pitch contour with reference to song and the song to be matched in the beginning and ending time
The similarity of pitch contour in point;
Integrated unit, for merging the similarity of the pitch contour with the similarity of sequence of notes;
Output unit, the fusion results for obtaining the integrated unit are exported as melodic similarity.
The candidate block specially selects melodic similarity greater than the song of given threshold as candidate song, or according to
The song of the descending sequential selection setting number of melodic similarity is as candidate song.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910306677.0A CN109992694A (en) | 2019-04-17 | 2019-04-17 | A kind of music intelligent recommendation method and system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910306677.0A CN109992694A (en) | 2019-04-17 | 2019-04-17 | A kind of music intelligent recommendation method and system |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109992694A true CN109992694A (en) | 2019-07-09 |
Family
ID=67133897
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910306677.0A Pending CN109992694A (en) | 2019-04-17 | 2019-04-17 | A kind of music intelligent recommendation method and system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109992694A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110598040A (en) * | 2019-09-17 | 2019-12-20 | 上海喜马拉雅科技有限公司 | Album recalling method, device, equipment and storage medium |
CN113032616A (en) * | 2021-03-19 | 2021-06-25 | 腾讯音乐娱乐科技(深圳)有限公司 | Audio recommendation method and device, computer equipment and storage medium |
CN113658594A (en) * | 2021-08-16 | 2021-11-16 | 北京百度网讯科技有限公司 | Lyric recognition method, device, equipment, storage medium and product |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101105815A (en) * | 2007-09-06 | 2008-01-16 | 腾讯科技(深圳)有限公司 | Internet music file sequencing method, system and search method and search engine |
CN104978962A (en) * | 2014-04-14 | 2015-10-14 | 安徽科大讯飞信息科技股份有限公司 | Query by humming method and system |
CN105930429A (en) * | 2016-04-19 | 2016-09-07 | 乐视控股(北京)有限公司 | Music recommendation method and apparatus |
CN106610968A (en) * | 2015-10-21 | 2017-05-03 | 广州酷狗计算机科技有限公司 | Song menu list determination method and apparatus, and electronic device |
CN108629047A (en) * | 2018-05-14 | 2018-10-09 | 平安科技(深圳)有限公司 | A kind of song list generation method and terminal device |
CN109408665A (en) * | 2018-12-29 | 2019-03-01 | 咪咕音乐有限公司 | A kind of information recommendation method and device, storage medium |
-
2019
- 2019-04-17 CN CN201910306677.0A patent/CN109992694A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101105815A (en) * | 2007-09-06 | 2008-01-16 | 腾讯科技(深圳)有限公司 | Internet music file sequencing method, system and search method and search engine |
CN104978962A (en) * | 2014-04-14 | 2015-10-14 | 安徽科大讯飞信息科技股份有限公司 | Query by humming method and system |
CN106610968A (en) * | 2015-10-21 | 2017-05-03 | 广州酷狗计算机科技有限公司 | Song menu list determination method and apparatus, and electronic device |
CN105930429A (en) * | 2016-04-19 | 2016-09-07 | 乐视控股(北京)有限公司 | Music recommendation method and apparatus |
CN108629047A (en) * | 2018-05-14 | 2018-10-09 | 平安科技(深圳)有限公司 | A kind of song list generation method and terminal device |
CN109408665A (en) * | 2018-12-29 | 2019-03-01 | 咪咕音乐有限公司 | A kind of information recommendation method and device, storage medium |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110598040A (en) * | 2019-09-17 | 2019-12-20 | 上海喜马拉雅科技有限公司 | Album recalling method, device, equipment and storage medium |
CN113032616A (en) * | 2021-03-19 | 2021-06-25 | 腾讯音乐娱乐科技(深圳)有限公司 | Audio recommendation method and device, computer equipment and storage medium |
CN113032616B (en) * | 2021-03-19 | 2024-02-20 | 腾讯音乐娱乐科技(深圳)有限公司 | Audio recommendation method, device, computer equipment and storage medium |
CN113658594A (en) * | 2021-08-16 | 2021-11-16 | 北京百度网讯科技有限公司 | Lyric recognition method, device, equipment, storage medium and product |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Knees et al. | A survey of music similarity and recommendation from music context data | |
Song et al. | A survey of music recommendation systems and future perspectives | |
Koenigstein et al. | Yahoo! music recommendations: modeling music ratings with temporal dynamics and item taxonomy | |
Knees et al. | A music search engine built upon audio-based and web-based similarity measures | |
Bogdanov et al. | How much metadata do we need in music recommendation?: A subjective evaluation using preference sets | |
CN101877711B (en) | Social network establishment method and device, and community discovery method and device | |
US20070282860A1 (en) | Method and system for music information retrieval | |
CN109992694A (en) | A kind of music intelligent recommendation method and system | |
Schedl et al. | A music information system automatically generated via web content mining techniques | |
KR20120101233A (en) | Method for providing sentiment information and method and system for providing contents recommendation using sentiment information | |
CN105279289B (en) | Individualized music based on exponential damping window recommends sort method | |
Darshna | Music recommendation based on content and collaborative approach & reducing cold start problem | |
Knees et al. | Music retrieval and recommendation: A tutorial overview | |
Wu | Music personalized recommendation system based on hybrid filtration | |
Miotto et al. | A probabilistic model to combine tags and acoustic similarity for music retrieval | |
Schedl et al. | Assigning and Visualizing Music Genres by Web-based Co-Occurrence Analysis. | |
Bonnin et al. | A comparison of playlist generation strategies for music recommendation and a new baseline scheme | |
Guo et al. | A query by humming system based on locality sensitive hashing indexes | |
Monti et al. | An ensemble approach of recurrent neural networks using pre-trained embeddings for playlist completion | |
Fields et al. | Using song social tags and topic models to describe and compare playlists | |
Tatli et al. | A tag-based hybrid music recommendation system using semantic relations and multi-domain information | |
Xiao et al. | Fast Hamming Space Search for Audio Fingerprinting Systems. | |
CN106021430A (en) | Full-text retrieval matching method and system based on Lucence custom lexicon | |
Schindler | Multi-modal music information retrieval: augmenting audio-analysis with visual computing for improved music video analysis | |
Su et al. | Semantic content-based music retrieval using audio and fuzzy-music-sense features |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190709 |