CN108920577A - Television set intelligently recommended method - Google Patents
Television set intelligently recommended method Download PDFInfo
- Publication number
- CN108920577A CN108920577A CN201810657522.7A CN201810657522A CN108920577A CN 108920577 A CN108920577 A CN 108920577A CN 201810657522 A CN201810657522 A CN 201810657522A CN 108920577 A CN108920577 A CN 108920577A
- Authority
- CN
- China
- Prior art keywords
- user
- resource
- project
- scoring
- information
- 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
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Two-Way Televisions, Distribution Of Moving Picture Or The Like (AREA)
Abstract
The invention discloses a kind of television set intelligently recommended methods, it is related to ntelligent television technolog field, the television set intelligently recommended method passes through Bayesian network synthetic user identity, user's use habit, the information such as user's expression, for the customized recommendation of user, more factors can be integrated than Bayesian network with traditional recommended method, so that the reasoning results are more nearly the actual wishes of user;On the other hand, compared with internet program resource, TV programme number amount and type are relatively fixed, user's evaluation information is relatively fewer and is difficult to collect quantity, it therefore is infeasible using single recommended method, the present invention is recommended for the different characteristics of Internet resources and TV station's resource using the cascade of Bayesian network and collaborative filtering, real-time is higher, and is more suitable for television recommendations application scenarios.
Description
Technical field
The present invention relates to ntelligent television technolog field, in particular to a kind of television set intelligently recommended method.
Background technique
The development of Web TV and artificial intelligence technology makes it possible that TV set platform has the function of video recommendations.
Present Web TV is in addition to there are also network video resources for TV station's resource, and TV station's resource and internet program resource are in number of users
Amount is different in terms of user information, and suitable suggested design is also different, accordingly, it is considered to a kind of television program recommendations and
The recommended method that internet program resource recommendation combines is also necessary.
The patent of Patent No. CN200810217886 recommends TV programme using bayes method.Firstly, logical
It crosses user's history record and extracts viewing behavior to characterize the rating hobby of user, then, according to the characteristic present section of TV programme
The each attribute of purpose by viewing behavior and plays behavior prediction user watching generally to TV programme according to bayesian algorithm
Rate recommends TV programme to user accordingly.
Collaborative filtering recommending technology based on article has been introduced into TV by the patent of Patent No. CN201720708711
During program is recommended, in the invention, the similitude between article and article is calculated by article rating matrix first, is then utilized
User's history score data and article article similarity matrix calculate user to the preference of other articles.But be not difficult to find out, on
User's history score data used by the method stated is a bit very crucial.TV is the equipment of multi-user a kind of, current
Television equipment seldom needs user to log in, therefore the history score data for program obtained is sampled for same equipment not
It must be the data of sole user, so the confidence level of the user's history score data obtained in this total situation is very low.
Therefore, collaborative filtering is not particularly suited for the recommendation for traditional tv program.And for existing internet program resource
There are many marking, very complete, and therefore the guarantee of energy maximum probability, collaborative filtering recommending technology is drawn from the same user
Enter in the recommendation of network program resource and is avoided that the above problem.
The patent of Patent No. CN200910038899 uses the interactive TV program recommendation method based on collaborative filtering,
When user obtains the similar program of actual program using remote control, recommendation server line upper module is found and is worked as in similarity square matrix
Several maximum projects of preceding item similarity, and user is recommended as a result.Project-based collaborative filtering is focused on
The similitude between TV programme is analyzed, is conservative for recommendation phase.And user hobby itself is extensive, just so-called object
To birds of the same feather flock together, things of a kind come together, people of a mind fall into the same group, the different types of program that user has never seen it, may be still interested, if using protecting
The recommended method kept may allow user to miss some programs, thus using the collaborative filtering based on user be it is necessary to
's.
Existing television video recommended technology there are the problem of:Internet program resource and TV programme resource are not accounted for
The difference for the intelligent recommendation model training data set that can be obtained, proposes targeted video recommendations processing method.
Summary of the invention
Technical problem to be solved by the invention is to provide a kind of television set intelligently recommended methods, utilize Bayesian network and association
With filtering cascade, carried out tentatively recommending station synchronization and network video according to local user's information first with Bayesian network
Resource carries out network video resource recommendation using the collaborative filtering of cloud platform later, on the one hand improves advisory speed, and in addition one
Aspect has more suited the video recommendations application scenarios of TV platform.
To achieve the above object, the present invention provides technical solution below:
The television set intelligently recommended method includes the following steps:
Step S101:User's information is obtained, before using this system, user should pass through APP login account first, and
It uploads and uses user information, after user opens TV in this step, set-top box device acquires user images by front camera and obtains
User identity and expression information are taken, all information are all attached to UNIX timestamp;
Step S102:The quick recommendation based on Bayes is opened based on user's information;
Step S103:Judge whether institute's recommendation results are network video resource, if not then entering step S106;
Step S104:Collaborative filtering recommending then is carried out using system cloud platform if network video resource;
Step S105:Recommendation results are presented to the user in the form of network video the Resources list;
Step S106:It is directly entered corresponding station synchronization;
Step S107:More new database.
Quickly recommend that specific step is as follows based on Bayes:
Step S201:Two naive Bayesian networks of training are recorded using the step S101 user's information obtained and habit
Model, one is recommended for program category, another recommends for specific station synchronization;Program category is recommended program category
As root node, it is there are two state, respectively network video resource and station synchronization resource, by user identity, when operation
Between, as leaf node, training program category recommends Naive Bayes Classifier model, specific TV for resource tag and emotional information
Platform program is recommended using specific station synchronization as root node, its status number is equal to the number of programs of specific TV station, equally will
User identity, operating time, resource tag and emotional information are as leaf node, the specific station synchronization naive Bayesian point of training
Class device model;
Step S202:The naive Bayesian network model that program category based on step S201 training is recommended makes inferences,
Provide the probability that current time user wants viewing network video resource;Reading current time user information, current time,
Emotional information, carries out Bayesian inference, and available user wants the probability of viewing network video resource;
Step S203:If probability is greater than 0.5, return recommendation results are network video resource, otherwise enter step S204, if
Probability is greater than 0.5, i.e. user wants viewing network video resource, so recommendation network video resource to user, otherwise recommends electricity
Television stations program resource is to user;
Step S204:It is pushed away using the naive Bayesian network model that the specific TV station of training in step S201 is recommended
Reason obtains the probability that current time user wants each station synchronization of viewing;When user wants viewing station synchronization money
When source, it is also necessary to recommend specific station synchronization to user, need to use specific station synchronization at this time and recommend simple pattra leaves
This classifier.User's information, current time are read in, emotional information recommends naive Bayesian point using specific station synchronization
Class device carries out Bayesian inference, obtains the probability that user wants each specific station synchronization of viewing;
Step S205:It is station synchronization that the station synchronization for choosing maximum probability, which is used as and returns to recommendation results,;Known
Under the premise of user wants the probability of each specific station synchronization of viewing, maximum that of select probability recommends user.
Based on collaborative filtering, specific step is as follows:
Step S301:User's information that processing step S101 is obtained, establishes user's model, to number of users according to progress
Pretreatment, establishes user's model, to the first number of users according to pre-processing, establishes first user's model, to user
Viewing internet program resource scores, and establishes user-project appraisal matrix, and evaluations matrix browses record and scoring according to user
Record show that the matrix R, m that user-project appraisal matrix is expressed as a m × n are numbers of users, and n is item number, wherein rijTable
Show i-th of user to the score value of j-th of project, user-project appraisal matrix R is:
Step S302:Garbled data, which is concentrated, has the user of intersection with user, contained by existing network program resource data set
Content is huge, but the Internet resources that most users and user like seeing are not intersections, so, it is not necessary that it calculates
The similarity of all users and user in data set, it is only necessary to calculate the user for having intersection with user, be provided with internet program
Source has the user of intersection to the anti-computation of table lookup of user and user;
Step S303:Calculate the similarity of user and each user;It is commented according to user-project that step S301 is established
Valence matrix, structuring user's-project scoring vector, vector data is scoring of the user to the project, if user does not have project
It scores, then user is set as 0 to the scoring of the project, according to the use for having intersection with user filtered out in step S302
Family, the user-project for constructing corresponding user respectively score vector, calculate user in data set each user it is similar
Degree, the scoring of each user can regard the vector of n dimension project spatially as, if user does not score to project,
User is then set as 0 to the scoring of the project.Similarity between user passes through the cosine angle measurement between vector.If user i and
Scoring of the user j on n dimension project space is expressed as vector i and vector j, then the similitude between user i and user j
Sim (i, j) is:
Molecule is the inner product of two users' scoring vectors, and denominator is the product of two user vector moulds;
Step S304:K neighbours similar with user are found out, according to the height of user and the similarity of each user
The low user concentrated to data is ranked up, and chooses the K neighbours that the highest K user of similarity is user;
Step S305:Internet program resource is ranked up according to recommendation, recommends the net that user likes in neighbours
In network program resource, the recommendation of each internet program resource is calculated according to the how far of neighbours and user, according to recommendation
Degree is ranked up internet program resource, recommends user,
The calculation method for generating recommendation network program resource is as follows:
Wherein sim (i, j) indicates the similitude between user i and user j, Rj,dIndicate nearest-neighbors user j to project d
Scoring,WithThe average score for respectively indicating user i and user j is searched in the nearest-neighbors collection NBSi of user and is used
Family, and using the value of the similarity of the first user and the user found as weight, then by neighbor user to the network section
The difference of the scoring and all scorings of this neighbor user of mesh resource is weighted and averaged, and predicts the first use according to this method
Scoring of the people to internet program resource is not evaluated, then the higher program of selection prediction scoring recommends the first user.
It is using the beneficial effect of above technical scheme:The television set intelligently recommended method passes through Bayesian network synthetic user
Identity, user's use habit, the information such as user's expression are the customized recommendation of user, compare Bayes with traditional recommended method
Network can integrate more factors, so that the reasoning results are more nearly the actual wishes of user;On the other hand, with internet program
Resource is compared, and TV programme number amount and type are relatively fixed, user's evaluation information is relatively fewer and are difficult to collect quantity, therefore adopt
Be with single recommended method it is infeasible, the present invention utilizes Bayes for the different characteristics of Internet resources and TV station's resource
The cascade of network and collaborative filtering is recommended, and real-time is higher, and is more suitable for television recommendations application scenarios.
Detailed description of the invention
A specific embodiment of the invention is described in further detail with reference to the accompanying drawing.
Fig. 1 is the flow chart of television set intelligently recommended method of the present invention;
Fig. 2 is the primary proposed algorithm flow chart based on Bayes of television set intelligently recommended method of the present invention;
Fig. 3 is the second level proposed algorithm flow chart based on collaborative filtering of television set intelligently recommended method of the present invention.
Specific embodiment
The preferred embodiment for television set intelligently recommended method that the invention will now be described in detail with reference to the accompanying drawings.
Fig. 1, Fig. 2 and Fig. 3 show the specific embodiment of television set intelligently recommended method of the present invention:
As shown in Figure 1, the television set intelligently recommended method includes the following steps:
Step S101:User's information is obtained, before using this system, user should pass through APP login account first, and
It uploads and uses user information, after user opens TV in this step, set-top box device acquires user images by front camera and obtains
User identity and expression information are taken, all information are all attached to UNIX timestamp;
Step S102:The quick recommendation based on Bayes is opened based on user's information;
Step S103:Judge whether institute's recommendation results are network video resource, if not then entering step S106;
Step S104:Collaborative filtering recommending then is carried out using system cloud platform if network video resource;
Step S105:Recommendation results are presented to the user in the form of network video the Resources list;
Step S106:It is directly entered corresponding station synchronization;
Step S107:More new database.
As shown in Fig. 2, quickly recommending that specific step is as follows based on Bayes:
Step S201:Two naive Bayesian networks of training are recorded using the step S101 user's information obtained and habit
Model, one is recommended for program category, another recommends for specific station synchronization;Program category is recommended program category
As root node, it is there are two state, respectively network video resource and station synchronization resource, by user identity, when operation
Between, as leaf node, training program category recommends Naive Bayes Classifier model, specific TV for resource tag and emotional information
Platform program is recommended using specific station synchronization as root node, its status number is equal to the number of programs of specific TV station, equally will
User identity, operating time, resource tag and emotional information are as leaf node, the specific station synchronization naive Bayesian point of training
Class device model;
Step S202:The naive Bayesian network model that program category based on step S201 training is recommended makes inferences,
Provide the probability that current time user wants viewing network video resource;Reading current time user information, current time,
Emotional information, carries out Bayesian inference, and available user wants the probability of viewing network video resource;
Step S203:If probability is greater than 0.5, return recommendation results are network video resource, otherwise enter step S204, if
Probability is greater than 0.5, i.e. user wants viewing network video resource, so recommendation network video resource to user, otherwise recommends electricity
Television stations program resource is to user;
Step S204:It is pushed away using the naive Bayesian network model that the specific TV station of training in step S201 is recommended
Reason obtains the probability that current time user wants each station synchronization of viewing;When user wants viewing station synchronization money
When source, it is also necessary to recommend specific station synchronization to user, need to use specific station synchronization at this time and recommend simple pattra leaves
This classifier.User's information, current time are read in, emotional information recommends naive Bayesian point using specific station synchronization
Class device carries out Bayesian inference, obtains the probability that user wants each specific station synchronization of viewing;
Step S205:It is station synchronization that the station synchronization for choosing maximum probability, which is used as and returns to recommendation results,;Known
Under the premise of user wants the probability of each specific station synchronization of viewing, maximum that of select probability recommends user.
As shown in figure 3, based on collaborative filtering, specific step is as follows:
Step S301:User's information that processing step S101 is obtained, establishes user's model, to number of users according to progress
Pretreatment, establishes user's model, to the first number of users according to pre-processing, establishes first user's model, to user
Viewing internet program resource scores, and establishes user-project appraisal matrix, and evaluations matrix browses record and scoring according to user
Record show that the matrix R, m that user-project appraisal matrix is expressed as a m × n are numbers of users, and n is item number, wherein rijTable
Show i-th of user to the score value of j-th of project, user-project appraisal matrix R is:
Step S302:Garbled data, which is concentrated, has the user of intersection with user, contained by existing network program resource data set
Content is huge, but the Internet resources that most users and user like seeing are not intersections, so, it is not necessary that it calculates
The similarity of all users and user in data set, it is only necessary to calculate the user for having intersection with user, be provided with internet program
Source has the user of intersection to the anti-computation of table lookup of user and user;
Step S303:Calculate the similarity of user and each user;It is commented according to user-project that step S301 is established
Valence matrix, structuring user's-project scoring vector, vector data is scoring of the user to the project, if user does not have project
It scores, then user is set as 0 to the scoring of the project, according to the use for having intersection with user filtered out in step S302
Family, the user-project for constructing corresponding user respectively score vector, calculate user in data set each user it is similar
Degree, the scoring of each user can regard the vector of n dimension project spatially as, if user does not score to project,
User is then set as 0 to the scoring of the project.Similarity between user passes through the cosine angle measurement between vector.If user i and
Scoring of the user j on n dimension project space is expressed as vector i and vector j, then the similitude between user i and user j
Sim (i, j) is:
Molecule is the inner product of two users' scoring vectors, and denominator is the product of two user vector moulds;
Step S304:K neighbours similar with user are found out, according to the height of user and the similarity of each user
The low user concentrated to data is ranked up, and chooses the K neighbours that the highest K user of similarity is user;
Step S305:Internet program resource is ranked up according to recommendation, recommends the net that user likes in neighbours
In network program resource, the recommendation of each internet program resource is calculated according to the how far of neighbours and user, according to recommendation
Degree is ranked up internet program resource, recommends user,
The calculation method for generating recommendation network program resource is as follows:
Wherein sim (i, j) indicates the similitude between user i and user j, Rj,dIndicate nearest-neighbors user j to project d
Scoring,WithThe average score for respectively indicating user i and user j is searched in the nearest-neighbors collection NBSi of user and is used
Family, and using the value of the similarity of the first user and the user found as weight, then by neighbor user to the network section
The difference of the scoring and all scorings of this neighbor user of mesh resource is weighted and averaged, and predicts the first use according to this method
Scoring of the people to internet program resource is not evaluated, then the higher program of selection prediction scoring recommends the first user.
The above are merely the preferred embodiment of the present invention, it is noted that for those of ordinary skill in the art,
Without departing from the concept of the premise of the invention, various modifications and improvements can be made, these belong to guarantor of the invention
Protect range.
Claims (3)
1. a kind of television set intelligently recommended method, it is characterised in that:The television set intelligently recommended method includes the following steps:
Step S101:User's information is obtained, before using this system, user should pass through APP login account first, and upload
With user information, after user opens TV in this step, set-top box device acquires user images by front camera and obtains use
Family identity and expression information, all information are all attached to UNIX timestamp;
Step S102:The quick recommendation based on Bayes is opened based on user's information;
Step S103:Judge whether institute's recommendation results are network video resource, if not then entering step S106;
Step S104:Collaborative filtering recommending then is carried out using system cloud platform if network video resource;
Step S105:Recommendation results are presented to the user in the form of network video the Resources list;
Step S106:It is directly entered corresponding station synchronization;
Step S107:More new database.
2. television set intelligently recommended method according to claim 1, it is characterised in that:The quick recommendation based on Bayes
Specific step is as follows:
Step S201:Two naive Bayesian network moulds of training are recorded using the step S101 user's information obtained and habit
Type, one is recommended for program category, another recommends for specific station synchronization;Program category is made in program category recommendation
For root node, it is there are two state, respectively network video resource and station synchronization resource, by user identity, the operating time,
As leaf node, training program category recommends Naive Bayes Classifier model, specific TV station for resource tag and emotional information
Program is recommended using specific station synchronization as root node, its status number is equal to the number of programs of specific TV station, will equally use
Family identity, operating time, resource tag and emotional information are as leaf node, the specific station synchronization Naive Bayes Classification of training
Device model;
Step S202:The naive Bayesian network model that program category based on step S201 training is recommended makes inferences, and provides
Current time user wants the probability of viewing network video resource;Read in current time user information, current time, mood
Information, carries out Bayesian inference, and available user wants the probability of viewing network video resource;
Step S203:If probability is greater than 0.5, return recommendation results are network video resource, S204 are otherwise entered step, if probability
Greater than 0.5, i.e., user wants viewing network video resource, so otherwise recommendation network video resource recommends TV station to user
Program resource is to user;
Step S204:It is made inferences using the naive Bayesian network model that the specific TV station of training in step S201 is recommended,
Obtain the probability that current time user wants each station synchronization of viewing;When user wants viewing station synchronization resource
When, it is also necessary to recommend specific station synchronization to user, needs to use specific station synchronization at this time and recommend naive Bayesian
Classifier.User's information, current time are read in, emotional information recommends Naive Bayes Classification using specific station synchronization
Device carries out Bayesian inference, obtains the probability that user wants each specific station synchronization of viewing;
Step S205:It is station synchronization that the station synchronization for choosing maximum probability, which is used as and returns to recommendation results,;In known users
Under the premise of the probability for wanting each specific station synchronization of viewing, maximum that of select probability recommends user.
3. television set intelligently recommended method according to claim 1, it is characterised in that:The specific step based on collaborative filtering
It is rapid as follows:
Step S301:User's information that processing step S101 is obtained, establishes user's model, to number of users according to being located in advance
Reason, establishes user's model, to the first number of users according to pre-processing, first user's model is established, to user's watched
Internet program resource scores, and establishes user-project appraisal matrix, evaluations matrix browses record according to user and scoring records
It obtains, the matrix R, m that user-project appraisal matrix is expressed as a m × n are numbers of users, and n is item number, wherein rijIndicate i-th
A user is to the score value of j-th of project, user-project appraisal matrix R:
Step S302:Garbled data concentrates the user for having intersection with user, content contained by existing network program resource data set
It is huge, but the Internet resources that most users and user like seeing are not intersections, so, it is not necessary that calculate data
Concentrate the similarity of all users and user, it is only necessary to calculate the user for having intersection with user, be arrived with internet program resource
The anti-computation of table lookup of user and user have the user of intersection;
Step S303:Calculate the similarity of user and each user;User-project appraisal the square established according to step S301
Battle array, structuring user's-project scoring vector, vector data is scoring of the user to the project, if user does not carry out project
Scoring, then be set as 0 to the scoring of the project for user, according to the user for having intersection with user filtered out in step S302,
User-project scoring the vector for constructing corresponding user respectively, calculates the similarity of each user in user and data set, often
The scoring of one user can regard the vector of n dimension project spatially as will use if user does not score to project
Family is set as 0 to the scoring of the project.Similarity between user passes through the cosine angle measurement between vector.If user i and user j exist
The scoring of n dimension project spatially is expressed as vector i and vector j, then the similitude sim (i, j) between user i and user j
For:
Molecule is the inner product of two users' scoring vectors, and denominator is the product of two user vector moulds;
Step S304:K neighbours similar with user are found out, it is right according to user and the height of the similarity of each user
User in data set is ranked up, and chooses the K neighbours that the highest K user of similarity is user;
Step S305:Internet program resource is ranked up according to recommendation, recommends the network section that user likes in neighbours
In mesh resource, the recommendation of each internet program resource is calculated according to the how far of neighbours and user, according to recommendation pair
Internet program resource is ranked up, and recommends user,
The calculation method for generating recommendation network program resource is as follows:
Wherein sim (i, j) indicates the similitude between user i and user j, Rj,dIndicate that nearest-neighbors user j comments project d
Point,WithThe average score for respectively indicating user i and user j searches user in the nearest-neighbors collection NBSi of user, and
Using the value of the similarity of the first user and the user found as weight, then by neighbor user to the internet program resource
Scoring and the differences of all scorings of this neighbor user be weighted and averaged, the first user is predicted to not according to this method
The scoring of internet program resource is evaluated, then the higher program of selection prediction scoring recommends the first user.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810657522.7A CN108920577A (en) | 2018-06-25 | 2018-06-25 | Television set intelligently recommended method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810657522.7A CN108920577A (en) | 2018-06-25 | 2018-06-25 | Television set intelligently recommended method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108920577A true CN108920577A (en) | 2018-11-30 |
Family
ID=64420847
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810657522.7A Pending CN108920577A (en) | 2018-06-25 | 2018-06-25 | Television set intelligently recommended method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108920577A (en) |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109672938A (en) * | 2019-01-07 | 2019-04-23 | 河北工业大学 | A kind of IPTV program commending method |
CN109873921A (en) * | 2019-02-14 | 2019-06-11 | 初速度(苏州)科技有限公司 | A kind of transmission method of image temporal stamp, analysis method and device |
CN110351580A (en) * | 2019-07-12 | 2019-10-18 | 四川长虹电器股份有限公司 | TV programme special recommendation method and system based on Non-negative Matrix Factorization |
CN110442977A (en) * | 2019-08-08 | 2019-11-12 | 广州华建工智慧科技有限公司 | Mobile terminal BIM model intelligent buffer method based on construction process network recommendation |
CN110572596A (en) * | 2019-09-11 | 2019-12-13 | 深圳创维-Rgb电子有限公司 | television and backlight control method and device thereof and readable storage medium |
CN110595012A (en) * | 2019-09-23 | 2019-12-20 | 海信(广东)空调有限公司 | Recommendation method and device for air conditioner operation mode |
CN111105298A (en) * | 2019-12-31 | 2020-05-05 | 杭州涂鸦信息技术有限公司 | Purchasing recommendation method and system based on intelligent scene of Internet of things |
CN111417017A (en) * | 2020-04-28 | 2020-07-14 | 安徽国广数字科技有限公司 | IPTV program recommendation method and system based on human body identification |
TWI710251B (en) * | 2019-03-29 | 2020-11-11 | 展雋創意股份有限公司 | One-touch-switch multi-user tv user interface and system thereof |
CN112740712A (en) * | 2019-10-10 | 2021-04-30 | 海信视像科技股份有限公司 | Receiving apparatus, server, content recommendation system, and content recommendation method |
US11006178B2 (en) | 2019-05-16 | 2021-05-11 | Ovomedia Creative Inc. | Multi-account personalized artificial intelligence TV interface and system thereof |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101287082A (en) * | 2008-05-16 | 2008-10-15 | 华东师范大学 | Collaborative filtered recommendation method introducing hotness degree weight of program |
CN102056018A (en) * | 2010-11-26 | 2011-05-11 | Tcl集团股份有限公司 | Method and system for providing TV guide and method for providing program-requesting information |
CN102841929A (en) * | 2012-07-19 | 2012-12-26 | 南京邮电大学 | Recommending method integrating user and project rating and characteristic factors |
CN105868334A (en) * | 2016-03-28 | 2016-08-17 | 云南财经大学 | Personalized film recommendation method and system based on feature augmentation |
CN107454474A (en) * | 2017-08-17 | 2017-12-08 | 四川长虹电器股份有限公司 | A kind of television terminal program personalized recommendation method based on collaborative filtering |
-
2018
- 2018-06-25 CN CN201810657522.7A patent/CN108920577A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101287082A (en) * | 2008-05-16 | 2008-10-15 | 华东师范大学 | Collaborative filtered recommendation method introducing hotness degree weight of program |
CN102056018A (en) * | 2010-11-26 | 2011-05-11 | Tcl集团股份有限公司 | Method and system for providing TV guide and method for providing program-requesting information |
CN102841929A (en) * | 2012-07-19 | 2012-12-26 | 南京邮电大学 | Recommending method integrating user and project rating and characteristic factors |
CN105868334A (en) * | 2016-03-28 | 2016-08-17 | 云南财经大学 | Personalized film recommendation method and system based on feature augmentation |
CN107454474A (en) * | 2017-08-17 | 2017-12-08 | 四川长虹电器股份有限公司 | A kind of television terminal program personalized recommendation method based on collaborative filtering |
Cited By (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109672938A (en) * | 2019-01-07 | 2019-04-23 | 河北工业大学 | A kind of IPTV program commending method |
CN109873921A (en) * | 2019-02-14 | 2019-06-11 | 初速度(苏州)科技有限公司 | A kind of transmission method of image temporal stamp, analysis method and device |
TWI710251B (en) * | 2019-03-29 | 2020-11-11 | 展雋創意股份有限公司 | One-touch-switch multi-user tv user interface and system thereof |
US11006178B2 (en) | 2019-05-16 | 2021-05-11 | Ovomedia Creative Inc. | Multi-account personalized artificial intelligence TV interface and system thereof |
CN110351580B (en) * | 2019-07-12 | 2021-07-13 | 四川长虹电器股份有限公司 | Television program topic recommendation method and system based on non-negative matrix factorization |
CN110351580A (en) * | 2019-07-12 | 2019-10-18 | 四川长虹电器股份有限公司 | TV programme special recommendation method and system based on Non-negative Matrix Factorization |
CN110442977A (en) * | 2019-08-08 | 2019-11-12 | 广州华建工智慧科技有限公司 | Mobile terminal BIM model intelligent buffer method based on construction process network recommendation |
CN110442977B (en) * | 2019-08-08 | 2023-09-29 | 广州华建工智慧科技有限公司 | Mobile terminal BIM model intelligent caching method based on building construction procedure network recommendation |
CN110572596A (en) * | 2019-09-11 | 2019-12-13 | 深圳创维-Rgb电子有限公司 | television and backlight control method and device thereof and readable storage medium |
US11388363B2 (en) | 2019-09-11 | 2022-07-12 | Shenzhen Skyworth-Rgb Electronic Co., Ltd. | TV set and backlight control method, control device and readable storage medium of the TV set |
CN110595012A (en) * | 2019-09-23 | 2019-12-20 | 海信(广东)空调有限公司 | Recommendation method and device for air conditioner operation mode |
CN112740712A (en) * | 2019-10-10 | 2021-04-30 | 海信视像科技股份有限公司 | Receiving apparatus, server, content recommendation system, and content recommendation method |
CN111105298B (en) * | 2019-12-31 | 2023-09-26 | 杭州涂鸦信息技术有限公司 | Purchasing recommendation method and system based on intelligent scene of Internet of things |
CN111105298A (en) * | 2019-12-31 | 2020-05-05 | 杭州涂鸦信息技术有限公司 | Purchasing recommendation method and system based on intelligent scene of Internet of things |
CN111417017A (en) * | 2020-04-28 | 2020-07-14 | 安徽国广数字科技有限公司 | IPTV program recommendation method and system based on human body identification |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108920577A (en) | Television set intelligently recommended method | |
CN103365936A (en) | Video recommendation system and method thereof | |
CN110430471A (en) | It is a kind of based on the television recommendations method and system instantaneously calculated | |
KR101816113B1 (en) | Estimating and displaying social interest in time-based media | |
WO2017096832A1 (en) | Media data recommendation method and server | |
CN106339507B (en) | Streaming Media information push method and device | |
CN104935970A (en) | Method for recommending television content and television client | |
CN106547908A (en) | A kind of information-pushing method and system | |
Pouli et al. | Personalized multimedia content retrieval through relevance feedback techniques for enhanced user experience | |
CN108595493B (en) | Media content pushing method and device, storage medium and electronic device | |
CN112597395A (en) | Object recommendation method, device, equipment and storage medium | |
CN107894998A (en) | Video recommendation method and device | |
CN111597446B (en) | Content pushing method and device based on artificial intelligence, server and storage medium | |
CN110362753A (en) | A kind of personalized neural network recommendation method and system based on user concealed feedback | |
CN111954087B (en) | Method and device for intercepting images in video, storage medium and electronic equipment | |
Tai et al. | Content-based recommendation using machine learning | |
CN109874032B (en) | Program topic personalized recommendation system and method for smart television | |
Ashraf et al. | Personalized news recommendation based on multi-agent framework using social media preferences | |
Kofler et al. | Uploader intent for online video: typology, inference, and applications | |
CN108932245A (en) | Video searching method and device | |
Eom et al. | Improving image tag recommendation using favorite image context | |
CN111223014B (en) | Method and system for online generation of subdivision scene teaching courses from a large number of subdivision teaching contents | |
CN108804492B (en) | Method and device for recommending multimedia objects | |
Dai et al. | A real-time video recommendation system for live programs | |
Krishnamoorthy et al. | TV shows popularity and performance prediction using CNN algorithm |
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 | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20181130 |
|
WD01 | Invention patent application deemed withdrawn after publication |