CN110351580A - TV programme special recommendation method and system based on Non-negative Matrix Factorization - Google Patents
TV programme special recommendation method and system based on Non-negative Matrix Factorization Download PDFInfo
- Publication number
- CN110351580A CN110351580A CN201910631011.2A CN201910631011A CN110351580A CN 110351580 A CN110351580 A CN 110351580A CN 201910631011 A CN201910631011 A CN 201910631011A CN 110351580 A CN110351580 A CN 110351580A
- Authority
- CN
- China
- Prior art keywords
- label
- data
- user
- matrix
- thematic
- 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
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/20—Servers specifically adapted for the distribution of content, e.g. VOD servers; Operations thereof
- H04N21/23—Processing of content or additional data; Elementary server operations; Server middleware
- H04N21/235—Processing of additional data, e.g. scrambling of additional data or processing content descriptors
- H04N21/2355—Processing of additional data, e.g. scrambling of additional data or processing content descriptors involving reformatting operations of additional data, e.g. HTML pages
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/20—Servers specifically adapted for the distribution of content, e.g. VOD servers; Operations thereof
- H04N21/25—Management operations performed by the server for facilitating the content distribution or administrating data related to end-users or client devices, e.g. end-user or client device authentication, learning user preferences for recommending movies
- H04N21/251—Learning process for intelligent management, e.g. learning user preferences for recommending movies
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/20—Servers specifically adapted for the distribution of content, e.g. VOD servers; Operations thereof
- H04N21/25—Management operations performed by the server for facilitating the content distribution or administrating data related to end-users or client devices, e.g. end-user or client device authentication, learning user preferences for recommending movies
- H04N21/258—Client or end-user data management, e.g. managing client capabilities, user preferences or demographics, processing of multiple end-users preferences to derive collaborative data
- H04N21/25866—Management of end-user data
- H04N21/25891—Management of end-user data being end-user preferences
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
- H04N21/43—Processing of content or additional data, e.g. demultiplexing additional data from a digital video stream; Elementary client operations, e.g. monitoring of home network or synchronising decoder's clock; Client middleware
- H04N21/435—Processing of additional data, e.g. decrypting of additional data, reconstructing software from modules extracted from the transport stream
- H04N21/4355—Processing of additional data, e.g. decrypting of additional data, reconstructing software from modules extracted from the transport stream involving reformatting operations of additional data, e.g. HTML pages on a television screen
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
- H04N21/45—Management operations performed by the client for facilitating the reception of or the interaction with the content or administrating data related to the end-user or to the client device itself, e.g. learning user preferences for recommending movies, resolving scheduling conflicts
- H04N21/4508—Management of client data or end-user data
- H04N21/4532—Management of client data or end-user data involving end-user characteristics, e.g. viewer profile, preferences
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
- H04N21/45—Management operations performed by the client for facilitating the reception of or the interaction with the content or administrating data related to the end-user or to the client device itself, e.g. learning user preferences for recommending movies, resolving scheduling conflicts
- H04N21/466—Learning process for intelligent management, e.g. learning user preferences for recommending movies
- H04N21/4668—Learning process for intelligent management, e.g. learning user preferences for recommending movies for recommending content, e.g. movies
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computer Graphics (AREA)
- Computing Systems (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The present invention relates to big data technical fields, present invention seek to address that existing television program recommendations cannot rationally cover the problem of user's portrait, it is proposed a kind of TV programme special recommendation method based on Non-negative Matrix Factorization, it include: the real-time label information for obtaining and being intended to when recommending TV programme to user, label information includes: user's representation data, program label data and thematic label data, and user's representation data is for indicating scoring of the user to each label；Non-negative Matrix Factorization model is constructed, label information is converted into full label matrix, decomposition dimension-reduction treatment is carried out to full label matrix according to Non-negative Matrix Factorization model and obtains probability matrix；User's representation data and thematic label data are rebuild according to the label information in probability matrix, and carries out the recommendation of TV programme special topic according to the user's representation data rebuild and thematic label data.The present invention realizes user's recommendation list reasonable layout on interest worlds by the merging of interest worlds.
Description
Technical field
The present invention relates to big data technical fields, relate in particular to a kind of TV programme special recommendation method and system.
Background technique
In the personalized recommendation of smart television, label in interest worlds has intersection, and this is not paid close attention to,
That is, certain several label is that there is the user of same interest psychology to be had a preference for, such as act class and swordsman's class, usually to action movie
Score it is high also score swordsman's class it is high.But in the algorithm for generating recommendation list, the main scoring according to user to label
It is ranked up, if user is to swordsman, movement, the scoring such as ancient costume is high, then the label that recommendation list program is covered is then most of
It is these.But score label placed in the middle on the upper side there is also user, for example, pursuing a goal with determination, take a risk, comedy etc..And due to the force of front
The program proportion of the subject matters such as chivalrous movement is excessive, leaves for and pursues a goal with determination, and takes a risk, the recommendation space of comedy etc. is minimum, so as to cause this
A possibility that part is pushed out is minimum.Label to cause recommendation list program to include cannot rationally cover user's portrait, make
Recommendation effect is obtained to be declined.
Non-negative Matrix Factorization (Nonnegative Matrix Factor), abbreviation NMF is by Lee and Seung in 1999
A kind of matrix disassembling method that year proposes on Nature Journal, it is nonnegative value (it is required that pure add that it, which keeps the institute after decomposing important,
The description of property), and at the same time realizing that nonlinear dimension about subtracts.NMF has been increasingly becoming signal processing, biomedical engineering, mould
Most popular one of multidimensional data handling implement in the research fields such as formula identification, computer vision and Image Engineering.
Summary of the invention
Present invention seek to address that existing television program recommendations cannot rationally cover the problem of user's portrait, propose that one kind is based on
The TV programme special recommendation method and system of Non-negative Matrix Factorization.
The technical proposal adopted by the invention to solve the above technical problems is that: the TV programme based on Non-negative Matrix Factorization are special
Inscribe recommended method, comprising the following steps:
Step 1. obtains the label information being intended to when recommending TV programme to user in real time, and the label information includes: user
Representation data, program label data and thematic label data, user's representation data is for indicating user to each label
Scoring, the program label data are for indicating TV program information, and the special topic label data is for indicating a kind of TV Festival
Mesh information；
Step 2. constructs Non-negative Matrix Factorization model, the label information is converted to full label matrix, according to described non-
Negative matrix decomposition model carries out decomposition dimension-reduction treatment to full label matrix and obtains probability matrix, and the probability matrix is according to special topic
The degree of association between label merges the label information of generation；
Step 3. rebuilds user's representation data and thematic label data according to the label information in the probability matrix,
And the recommendation of TV programme special topic is carried out according to the user's representation data rebuild and thematic label data.
Further, be the decomposition for realizing full label matrix, in step 2, it is described by Non-negative Matrix Factorization model to complete
Label matrix carries out decomposition dimension-reduction treatment and obtains probability matrix specifically:
The Non-negative Matrix Factorization model is trained using program label data as training data, user is drawn a portrait number
It is fitted according to thematic label data as fitting data, obtains full label matrix and decompose the probability matrix after dimension-reduction treatment.
It further, is the recommendation for realizing TV programme special topic, in step 3, the user that the basis is rebuild draws a portrait
Data and the recommendation of thematic label data progress TV programme special topic include:
The cosine similarity between the user's representation data rebuild and thematic label data is calculated, to the cosine phase
Like degree be ranked up from large to small, select cosine similarity ranking before N the corresponding TV programme special topic of thematic label data into
Row is recommended, and the N is the positive integer more than or equal to 1.
It further, is the calculating for realizing cosine similarity, the cosine similarity calculation formula are as follows:
In formula, x indicates that the corresponding vector mould of user's representation data is long, and y indicates that the corresponding vector mould of thematic label data is long,
T indicates transposition.
The present invention also proposes a kind of TV programme special recommendation system based on Non-negative Matrix Factorization, comprising:
Acquiring unit, for obtaining the label information being intended to when recommending TV programme to user, the label information packet in real time
Include: user's representation data, program label data and thematic label data, user's representation data is for indicating user to each
The scoring of label, the program label data are for indicating TV program information, and the special topic label data is for indicating a kind of
TV program information；
Non-negative Matrix Factorization model obtains probability matrix for carrying out decomposition dimension-reduction treatment to full label matrix, described complete
Label matrix is converted to by the label information, and the probability matrix is to be merged according to the degree of association between thematic label
The label information of generation；
Recommendation unit, for rebuilding user's representation data and special topic mark according to the label information in the probability matrix
Data are signed, and carry out the recommendation of TV programme special topic according to the user's representation data rebuild and thematic label data.
Further, it is described to full label matrix carry out decompose dimension-reduction treatment obtain probability matrix specifically:
The Non-negative Matrix Factorization model is trained using program label data as training data, user is drawn a portrait number
It is fitted according to thematic label data as fitting data, obtains full label matrix and carry out the probability square after decomposition dimension-reduction treatment
Battle array.
Further, the recommendation unit is also used to:
The cosine similarity between the user's representation data rebuild and thematic label data is calculated, to the cosine phase
Like degree be ranked up from large to small, select cosine similarity ranking before N the corresponding TV programme special topic of thematic label data into
Row is recommended, and the N is the positive integer more than or equal to 1.
Further, the cosine similarity calculation formula are as follows:
In formula, x indicates that the corresponding vector mould of user's representation data is long, and y indicates that the corresponding vector mould of thematic label data is long,
T indicates transposition.
The beneficial effects of the present invention are: the TV programme special recommendation method of the present invention based on Non-negative Matrix Factorization
And system, by being decomposed based on Non-negative Matrix Factorization model to full label matrix, drawing a portrait after obtaining dimensionality reduction comprising user
The probability matrix of data and thematic label data, is optimized label, by the merging of interest worlds, finally realizes user
Recommendation list reasonable layout on interest worlds.
Detailed description of the invention
Fig. 1 is Non-negative Matrix Factorization schematic diagram；
Fig. 2 is the flow diagram of the TV programme special recommendation method of the present invention based on Non-negative Matrix Factorization；
Fig. 3 is the structural schematic diagram of the TV programme special recommendation system of the present invention based on Non-negative Matrix Factorization.
Specific embodiment
Embodiments of the present invention are described in detail below in conjunction with attached drawing.
The various dimension reduction methods of comparative analysis, in high dimensional data dimensionality reduction, because based on the dimension reduction method of distance in higher-dimension
Arest neighbors and farthest neighbour can be shown in data in most cases almost second countable from dimensionality reduction effect is poor, so based on distance
Dimension reduction method is no longer applicable in.NMF Non-negative Matrix Factorization is then to carry out dimensionality reduction, the mark of TV programme based on the degree of association between feature
Label data belong to high dimensional data, and by the degree of association between analyzing tags come the friendship between analyzing tags in interest worlds
Collection, and then decide whether to merge dimensionality reduction, so, the present invention carries out label compression using NMF Non-negative Matrix Factorization.
When carrying out compression dimensionality reduction using label of the Non-negative Matrix Factorization to special recommendation, dimension reduc-ing principle as shown in Figure 1,
Label matrix in special recommendation before dimensionality reduction is full label matrix V, and full label matrix V is F × N rank matrix, and matrix W is each
All labels scoring of a user is rating matrix, and rating matrix W is F × kth moment battle array, and H is K × N rank probability matrix, probability
Matrix H is used to reflect the label data after merging according to the intersection between label in interest worlds.
TV programme special recommendation method of the present invention based on Non-negative Matrix Factorization, as shown in Figure 1, including following
Step: step S1. obtains the label information being intended to when recommending TV programme to user in real time, and the label information includes: user's picture
As data, program label data and thematic label data, user's representation data is for indicating that user comments each label
Point, the program label data are for indicating TV program information, and the special topic label data is for indicating a kind of TV programme
Information；Step S2. constructs Non-negative Matrix Factorization model, the label information is converted to full label matrix, by described non-negative
Matrix decomposition model carries out decomposition dimension-reduction treatment to full label matrix and obtains probability matrix, and the probability matrix is to be marked according to special topic
The degree of association between label merges the label information of generation；Step S3. according to the label information in the probability matrix again
User's representation data and thematic label data are constructed, and is carried out according to the user's representation data rebuild and thematic label data
The recommendation of TV programme special topic.
Firstly, routine according to user to the scoring of TV programme label to carry out television program recommendations on the basis of, obtain
The label information being intended to when recommending TV programme to user is taken, the label information includes: user's representation data, program label data
With thematic label data, wherein for indicating scoring of the user to each label, program label data are used for user's representation data
Indicate TV program information, for example, Journey to the West, A Dream of Red Mansions etc., thematic label data is for indicating a kind of TV program information, example
Such as, swordsman's class, ancient costume class, love class, family's class etc., user's representation data, program label data and thematic label data three
Label dimension be that dynamic increases and real-time update, it is assumed that the dimension of three is respectively A, B, C, then its union complete or collected works is taken to make
For label dimension, i.e. A ∪ B ∪ C, and all labels are subjected to digital coding according to the size of union dimension.
Then, Non-negative Matrix Factorization model is constructed, it is complete that all label informations after progress digital coding are converted to array
Label matrix, and be input in the Non-negative Matrix Factorization model of building and decomposition dimensionality reduction is carried out to full label matrix, specifically, just
It is that the intersection between label on interest worlds is obtained, and melted according to intersection according to the degree of association analyzed between each label
Close dimensionality reduction, obtain corresponding rating matrix and probability matrix, wherein probability matrix be according to the degree of association between thematic label into
Row merges the label information generated.
Wherein, carry out decomposing dimension-reduction treatment that obtain probability matrix specific to full label matrix by Non-negative Matrix Factorization model
It can be with are as follows: the Non-negative Matrix Factorization model is trained using program label data as training data, user is drawn a portrait number
It is fitted according to thematic label data as fitting data, obtains full label matrix and decompose the probability matrix after dimension-reduction treatment.
Finally, carrying out pushing away for TV programme special topic with thematic label data according to user's representation data in probability matrix
It recommends, specifically, the cosine similarity between the user's representation data rebuild and thematic label data can be calculated, to described
Cosine similarity is ranked up from large to small, selects the corresponding TV programme of thematic label data of N before cosine similarity ranking
Special topic is recommended, and the N is the positive integer more than or equal to 1.Wherein, the thematic label data rebuild then contains newly
Thematic label data, for example, acrobatic fighting class, emotion class etc., wherein acrobatic fighting class then contains swordsman's class and ancient costume class, emotion class
Then contain love class and family's class.And then label is optimized, it is final to realize that user pushes away by the merging of interest worlds
Recommend list reasonable layout on interest worlds.
Wherein, the calculation formula of cosine similarity can be with are as follows:
In formula, x indicates that the corresponding vector mould of user's representation data is long, and y indicates that the corresponding vector mould of thematic label data is long,
T indicates transposition.
Based on the above-mentioned technical proposal, the present invention also proposes a kind of TV programme special recommendation system based on Non-negative Matrix Factorization
System, as shown in Figure 3, comprising:
Acquiring unit, for obtaining the label information being intended to when recommending TV programme to user, the label information packet in real time
Include: user's representation data, program label data and thematic label data, user's representation data is for indicating user to each
The scoring of label, the program label data are for indicating TV program information, and the special topic label data is for indicating a kind of
TV program information；
Non-negative Matrix Factorization model obtains probability matrix for carrying out decomposition dimension-reduction treatment to full label matrix, described complete
Label matrix is converted to by the label information, and the probability matrix is to be merged according to the degree of association between thematic label
The label information of generation；
Recommendation unit, for rebuilding user's representation data and special topic mark according to the label information in the probability matrix
Data are signed, and carry out the recommendation of TV programme special topic according to the user's representation data rebuild and thematic label data.
Optionally, it is described to full label matrix carry out decompose dimension-reduction treatment obtain probability matrix specifically:
The Non-negative Matrix Factorization model is trained using program label data as training data, user is drawn a portrait number
It is fitted according to thematic label data as fitting data, obtains full label matrix and carry out the probability square after decomposition dimension-reduction treatment
Battle array.
The optional recommendation unit is also used to:
The cosine similarity between the user's representation data rebuild and thematic label data is calculated, to the cosine phase
Like degree be ranked up from large to small, select cosine similarity ranking before N the corresponding TV programme special topic of thematic label data into
Row is recommended, and the N is the positive integer more than or equal to 1.
Optionally, the cosine similarity calculation formula are as follows:
In formula, x indicates that the corresponding vector mould of user's representation data is long, and y indicates that the corresponding vector mould of thematic label data is long,
T indicates transposition.
It is appreciated that since the TV programme special recommendation system of the present invention based on Non-negative Matrix Factorization is to be used for
The system for realizing the TV programme special recommendation method based on Non-negative Matrix Factorization, for disclosed system, due to
It is corresponding with disclosed method, so description is relatively simple, related place illustrates referring to the part of method.Due to upper
State the TV programme special recommendation method based on Non-negative Matrix Factorization be able to solve solve existing television program recommendations cannot be reasonable
The problem of covering user's portrait, therefore, the system for realizing the above-mentioned TV programme special recommendation method based on Non-negative Matrix Factorization
Equally it is able to solve the problem of existing television program recommendations cannot rationally cover user's portrait.
Claims (8)
1. the TV programme special recommendation method based on Non-negative Matrix Factorization, which comprises the following steps:
Step 1. obtains the label information being intended to when recommending TV programme to user in real time, and the label information includes: user's portrait
Data, program label data and thematic label data, user's representation data for indicating scoring of the user to each label,
The program label data are for indicating TV program information, and the special topic label data is for indicating a kind of TV programme letter
Breath；
Step 2. constructs Non-negative Matrix Factorization model, and the label information is converted to full label matrix, passes through the non-negative square
Battle array decomposition model carries out decomposition dimension-reduction treatment to full label matrix and obtains probability matrix, and the probability matrix is according to thematic label
Between the degree of association merge the label information of generation；
Step 3. rebuilds user's representation data and thematic label data, and root according to the label information in the probability matrix
The recommendation of TV programme special topic is carried out according to the user's representation data rebuild and thematic label data.
2. the TV programme special recommendation method based on Non-negative Matrix Factorization as described in claim 1, which is characterized in that step
In 2, it is described by Non-negative Matrix Factorization model to full label matrix carry out decompose dimension-reduction treatment obtain probability matrix specifically:
The Non-negative Matrix Factorization model is trained using program label data as training data, by user's representation data and
Thematic label data is fitted as fitting data, is obtained full label matrix and is decomposed the probability matrix after dimension-reduction treatment.
3. the TV programme special recommendation method based on Non-negative Matrix Factorization as described in claim 1, which is characterized in that step
In 3, the recommendation that the user's representation data and thematic label data that the basis is rebuild carry out TV programme special topic includes:
The cosine similarity between the user's representation data rebuild and thematic label data is calculated, to the cosine similarity
It is ranked up from large to small, the corresponding TV programme special topic of thematic label data of N is pushed away before selection cosine similarity ranking
It recommends, the N is the positive integer more than or equal to 1.
4. the TV programme special recommendation method based on Non-negative Matrix Factorization as claimed in claim 3, which is characterized in that described
Cosine similarity calculation formula are as follows:
In formula, x indicates that the corresponding vector mould of user's representation data is long, and y indicates that the corresponding vector mould of thematic label data is long, T table
Show transposition.
5. the TV programme special recommendation system based on Non-negative Matrix Factorization characterized by comprising
Acquiring unit, for obtaining the label information being intended to when recommending TV programme to user in real time, the label information includes: use
Family representation data, program label data and thematic label data, user's representation data is for indicating user to each label
Scoring, the program label data are for indicating TV program information, and the special topic label data is for indicating a kind of TV
Programme information；
Non-negative Matrix Factorization model obtains probability matrix, the full label for carrying out decomposition dimension-reduction treatment to full label matrix
Matrix is converted to by the label information, and the probability matrix is to merge generation according to the degree of association between thematic label
Label information；
Recommendation unit, for rebuilding user's representation data and thematic number of tags according to the label information in the probability matrix
According to, and according to the recommendation of the user's representation data rebuild and thematic label data progress TV programme special topic.
6. the TV programme special recommendation system based on Non-negative Matrix Factorization as claimed in claim 5, which is characterized in that described
Decomposition dimension-reduction treatment is carried out to full label matrix and obtains probability matrix specifically:
The Non-negative Matrix Factorization model is trained using program label data as training data, by user's representation data and
Thematic label data is fitted as fitting data, is obtained full label matrix and is carried out the probability matrix after decomposition dimension-reduction treatment.
7. the TV programme special recommendation system based on Non-negative Matrix Factorization as claimed in claim 5, which is characterized in that described
Recommendation unit is also used to:
The cosine similarity between the user's representation data rebuild and thematic label data is calculated, to the cosine similarity
It is ranked up from large to small, the corresponding TV programme special topic of thematic label data of N is pushed away before selection cosine similarity ranking
It recommends, the N is the positive integer more than or equal to 1.
8. the TV programme special recommendation system based on Non-negative Matrix Factorization as claimed in claim 5, which is characterized in that described
Cosine similarity calculation formula are as follows:
In formula, x indicates that the corresponding vector mould of user's representation data is long, and y indicates that the corresponding vector mould of thematic label data is long, T table
Show transposition.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910631011.2A CN110351580B (en) | 2019-07-12 | 2019-07-12 | Television program topic recommendation method and system based on non-negative matrix factorization |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910631011.2A CN110351580B (en) | 2019-07-12 | 2019-07-12 | Television program topic recommendation method and system based on non-negative matrix factorization |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110351580A true CN110351580A (en) | 2019-10-18 |
| CN110351580B CN110351580B (en) | 2021-07-13 |
Family
ID=68176122
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910631011.2A Active CN110351580B (en) | 2019-07-12 | 2019-07-12 | Television program topic recommendation method and system based on non-negative matrix factorization |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110351580B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111507293A (en) * | 2020-04-22 | 2020-08-07 | 浙江工业大学 | Signal classification method based on graph convolution model |
| CN112637684A (en) * | 2020-12-25 | 2021-04-09 | 四川长虹电器股份有限公司 | Method for detecting user portrait label at smart television terminal |
| WO2022141862A1 (en) * | 2020-12-30 | 2022-07-07 | 平安科技（深圳）有限公司 | User portrait generation method and apparatus, computer device and storage medium |
| CN116017070A (en) * | 2022-12-01 | 2023-04-25 | 四川长虹电器股份有限公司 | Method for improving clicking rate of television homepage based on operation strategy |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070199040A1 (en) * | 2006-02-23 | 2007-08-23 | Lawrence Kates | Multi-channel parallel digital video recorder |
| CN101287082A (en) * | 2008-05-16 | 2008-10-15 | 华东师范大学 | Collaborative filtered recommendation method introducing hotness degree weight of program |
| CN101540874A (en) * | 2009-04-23 | 2009-09-23 | 中山大学 | Interactive TV program recommendation method based on collaborative filtration |
| US20100287588A1 (en) * | 2009-05-07 | 2010-11-11 | Stuart Cox | Method and apparatus for providing enhanced electronic program guide with personalized selection of broadcast content using affinities data and user preferences |
| CN103888852A (en) * | 2014-03-24 | 2014-06-25 | 清华大学 | Video recommendation method and device for social television |
| CN105282616A (en) * | 2015-11-13 | 2016-01-27 | 云南大学 | Combined collaborative filtering IPTV program recommendation method |
| CN105430505A (en) * | 2015-11-13 | 2016-03-23 | 云南大学 | IPTV program recommending method based on combined strategy |
| CN106940801A (en) * | 2016-01-04 | 2017-07-11 | 中国科学院声学研究所 | A kind of deeply for Wide Area Network learns commending system and method |
| CN108419134A (en) * | 2018-02-05 | 2018-08-17 | 华南理工大学 | The recommendation of the channels method merged with group current behavior based on individual history |
| CN108491527A (en) * | 2018-03-28 | 2018-09-04 | 百度在线网络技术（北京）有限公司 | Information recommendation method, device and electronic equipment |
| CN108920577A (en) * | 2018-06-25 | 2018-11-30 | 西北工业大学 | Television set intelligently recommended method |
| CN109597899A (en) * | 2018-09-26 | 2019-04-09 | 中国传媒大学 | The optimization method of media personalized recommendation system |
-
2019
- 2019-07-12 CN CN201910631011.2A patent/CN110351580B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070199040A1 (en) * | 2006-02-23 | 2007-08-23 | Lawrence Kates | Multi-channel parallel digital video recorder |
| CN101287082A (en) * | 2008-05-16 | 2008-10-15 | 华东师范大学 | Collaborative filtered recommendation method introducing hotness degree weight of program |
| CN101540874A (en) * | 2009-04-23 | 2009-09-23 | 中山大学 | Interactive TV program recommendation method based on collaborative filtration |
| US20100287588A1 (en) * | 2009-05-07 | 2010-11-11 | Stuart Cox | Method and apparatus for providing enhanced electronic program guide with personalized selection of broadcast content using affinities data and user preferences |
| CN103888852A (en) * | 2014-03-24 | 2014-06-25 | 清华大学 | Video recommendation method and device for social television |
| CN105282616A (en) * | 2015-11-13 | 2016-01-27 | 云南大学 | Combined collaborative filtering IPTV program recommendation method |
| CN105430505A (en) * | 2015-11-13 | 2016-03-23 | 云南大学 | IPTV program recommending method based on combined strategy |
| CN106940801A (en) * | 2016-01-04 | 2017-07-11 | 中国科学院声学研究所 | A kind of deeply for Wide Area Network learns commending system and method |
| CN108419134A (en) * | 2018-02-05 | 2018-08-17 | 华南理工大学 | The recommendation of the channels method merged with group current behavior based on individual history |
| CN108491527A (en) * | 2018-03-28 | 2018-09-04 | 百度在线网络技术（北京）有限公司 | Information recommendation method, device and electronic equipment |
| CN108920577A (en) * | 2018-06-25 | 2018-11-30 | 西北工业大学 | Television set intelligently recommended method |
| CN109597899A (en) * | 2018-09-26 | 2019-04-09 | 中国传媒大学 | The optimization method of media personalized recommendation system |
Non-Patent Citations (2)
| Title |
|---|
| 宋艳: "《基于用户行为的社交电视推荐算法研究》", 《中国优秀硕士学位论文全文数据库》 * |
| 蓝琪: "《基于聚类与协同过滤组合算法的数字电视推荐***设计》", 《有线电视技术》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111507293A (en) * | 2020-04-22 | 2020-08-07 | 浙江工业大学 | Signal classification method based on graph convolution model |
| CN112637684A (en) * | 2020-12-25 | 2021-04-09 | 四川长虹电器股份有限公司 | Method for detecting user portrait label at smart television terminal |
| WO2022141862A1 (en) * | 2020-12-30 | 2022-07-07 | 平安科技（深圳）有限公司 | User portrait generation method and apparatus, computer device and storage medium |
| CN116017070A (en) * | 2022-12-01 | 2023-04-25 | 四川长虹电器股份有限公司 | Method for improving clicking rate of television homepage based on operation strategy |
| CN116017070B (en) * | 2022-12-01 | 2024-04-12 | 四川长虹电器股份有限公司 | Method for improving clicking rate of television homepage based on operation strategy |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110351580B (en) | 2021-07-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110351580A (en) | TV programme special recommendation method and system based on Non-negative Matrix Factorization | |
| CN109697629B (en) | Product data pushing method and device, storage medium and computer equipment | |
| CN103365936A (en) | Video recommendation system and method thereof | |
| CN108764141B (en) | Game scene description method, device, equipment and storage medium thereof | |
| US8577962B2 (en) | Server apparatus, client apparatus, content recommendation method, and program | |
| JP5212610B2 (en) | Representative image or representative image group display system, method and program thereof, and representative image or representative image group selection system, method and program thereof | |
| CN111143613B (en) | Method, system, electronic device and storage medium for selecting video cover | |
| US20230353828A1 (en) | Model-based data processing method and apparatus | |
| CN106156693A (en) | The robust error correction method represented based on multi-model for facial recognition | |
| CN103207888A (en) | Product search device and product search method | |
| CN101668176A (en) | Multimedia content-on-demand and sharing method based on social interaction graph | |
| Gottschall et al. | Patterns of characterization in folktales across geographic regions and levels of cultural complexity: Literature as a neglected source of quantitative data | |
| CN106776716A (en) | A kind of intelligent Matching marketing consultant and the method and apparatus of user | |
| KR101639987B1 (en) | Method and Apparatus for Movies Recommendation by Using Hybrid Filtering Based | |
| CN107943990A (en) | More video summarization methods of archetypal analysis technology based on Weight | |
| Verleysen et al. | Video dataset of human demonstrations of folding clothing for robotic folding | |
| Burroughs | Statistics and baseball fandom: Sabermetric infrastructure of expertise | |
| CN113761253A (en) | Video tag determination method, device, equipment and storage medium | |
| CN112685596B (en) | Video recommendation method and device, terminal and storage medium | |
| CN109710802B (en) | Video classification method and device | |
| CN115965439A (en) | Data recall method, data recall device, electronic equipment and storage medium | |
| CN115686199A (en) | Method and device for generating group eye movement track, computing equipment and storage medium | |
| US20170351747A1 (en) | Data analysis system, data analysis method, and data analysis program | |
| CN113971595A (en) | Commodity recommendation method, system and equipment | |
| Kang et al. | Instance segmentation method of user interface component of games |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |