CN103546773A - Television program recommendation method and system - Google Patents
Television program recommendation method and system Download PDFInfo
- Publication number
- CN103546773A CN103546773A CN201310355368.5A CN201310355368A CN103546773A CN 103546773 A CN103546773 A CN 103546773A CN 201310355368 A CN201310355368 A CN 201310355368A CN 103546773 A CN103546773 A CN 103546773A
- Authority
- CN
- China
- Prior art keywords
- user
- program
- recommendation
- channels
- data
- 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
Images
Landscapes
- Two-Way Televisions, Distribution Of Moving Picture Or The Like (AREA)
Abstract
The invention discloses a television program recommendation method and system. The method includes the steps of A, a client acquires user channel changing data and transmits the same to a server; B, the server performs log analysis on the user channel changing data to obtain user behavior data of television program watching, and performs hot channel, user preference and similarity analysis according to the user channel changing data to obtain the channel recommendation result; C, the server performs feature weight analysis on the user behavior data and performs similarity analysis on EPG (electronic program guide) program data analysis to obtain the program recommendation result; D, synthesizing the channel recommendation result and the program recommendation result to adjust the television channel list. By the method, the programs, audience wanted to watch the most, in the live programs can be given, user preference analysis accuracy can be improved greatly, and quality of program recommendation is improved.
Description
Technical field
The present invention relates to intelligent television, particularly a kind of recommend method of TV programme and system thereof.
Background technology
The fast development of Internet has been brought the mankind into information-intensive society and the age of Internet economy, and enterprise development and personal lifestyle have all been produced to deep effect.The type of merchandize providing in the virtual environment Xia, of ecommerce businessman and quantity are more and more, and in the face of " ocean " of merchandise news, consumer is difficult to fast and effeciently pick out needed commodity.Under this background, commending system (Recommender Systems) arises at the historic moment, and it is the feature according to user, such as hobby, recommendation meets the object of customer requirements, also claims personalized recommendation system (Personalized Recommender Systems).
For TV programme, have equally such problem, increasing channel and more and more abundanter programme content allow televiewer be difficult to select.Traditional navigation can not meet the demand of TV user, and all there has been electric program menu (EPG, ElectronicProgramGuid) in TV station now, can obtain easily the content that TV station will broadcast within following a period of time.For facilitate user can fast finding to the program of liking, produced live telecast commending system, it is according to hobby and the corresponding TV programme of custom recommendation of different TV user colony.Existing television recommendations is divided into non-personalized recommendation and personalized recommendation.Non-personalized recommendation is exactly that the similar television channel commending of television channel watched by hot broadcast television channel at that time or with user is to user.This way of recommendation recommends to have certain effect according to the degree of current television program hot broadcast, but might not meet user's demand; And because the programme variety of current TV station is very abundant, the associated recommendation of simple television channel can not guarantee the correlation in specific time period TV programme, so with respect to personalized recommendation, non-personalized recommendation inferior position is larger.Personalized recommendation is by recording the history of watching of unique user, and this user's the behavior of watching is analyzed, thereby obtains user's hobby, and then user is recommended targetedly, compare non-personalized recommendation, this recommendation Objective is strong, and accuracy promotes greatly.
But, in existing live telecast, personalized recommendation system is mostly the recommendation for television channel, diving tower behavior while user's behavior conventionally can only recording user being watched to TV due to terminal (as TV set-top box), so most of live personalized recommendation is, for user, the behavior of watching of channel is directly recommended to channel.Due to channel program abundant species, the program category of different periods differs equally, and recommendation effect can not be guaranteed.And for the recommendation of program, recording user zapping in the time period all programme contents of watching analyze, otherness due to TV user colony, recommending the program that active user watches may be the type that other users like, not necessarily active user wants to watch, so user behavior record has the behavior of the non-autonomous generation of user, the outcome quality that such behavior is recommended out is not high yet.
Summary of the invention
In view of above-mentioned the deficiencies in the prior art part, the object of the present invention is to provide a kind of recommend method and system thereof of TV programme, to user's diving tower, behavior is analyzed, and synthesizer channel is recommended and program commending, improve the accuracy that user preferences degree is analyzed, guaranteed programs recommended quality.。
In order to achieve the above object, the present invention has taked following technical scheme:
A recommend method for TV programme, it comprises:
A, client are obtained user's diving tower data and are transferred to service end;
B, service end are carried out log analysis to described user's diving tower data and are obtained user behavior data, according to user's diving tower data, carry out the recommendation of the channels result that VHI, user preferences and similarity analysis obtain;
C, service end carry out feature weight analysis and EPG program data are carried out to similarity analysis obtaining program commending result to described user behavior data;
D, comprehensive described recommendation of the channels result and program commending result are adjusted television channel lists.
In the recommend method of described TV programme, in described step B, service end is carried out log analysis to diving tower data and is obtained user behavior data and specifically comprise:
B1, according to diving tower data analysis user to program watch frequency and viewing time, generate user behavior data;
B2, store described user behavior data.
In the recommend method of described TV programme, in described step B, the recommendation of the channels result of carrying out VHI, user preferences and similarity analysis acquisition according to user's diving tower data specifically comprises:
B21, according to described user's diving tower, the channel that number is maximum is watched in data statistics, generates popular channel bank name; Statistics unique user is watched duration the sequence of each channel in each period, generate user preferred channels rank; The behavior of watching of counting user is carried out cluster analysis to user, generates similar users table;
B22, according to user preferred channels rank correction similar users table, then carry out user collaborative and filter to generate collaborative filtering recommendation of the channels table;
B23, after being merged, filter, sorted, VHI rank, user preferred channels rank and collaborative filtering recommendation of the channels table generate recommendation of the channels result; According to VHI rank and user preferred channels rank, be accustomed to modeling, generate user channel custom model.
In the recommend method of described TV programme, described step C specifically comprises:
C1, EPG program data is carried out to feature extraction and feature weight analysis, generating feature weight database; To user behavior data according to time attenuation function, in conjunction with feature weight Database user characteristics model;
C2, from EPG program data, extract the content information of program, according to described content information and feature weight database, carry out modeling, generate programs feature model;
C3, described EPG program data is carried out to correlation calculations, generate program Relational database;
C4, obtain real-time hot broadcast channel and by temperature sequence, in conjunction with EPG program data, extract hot broadcast program, described programs feature model mate with user characteristics model and obtain degree of correlation seniority among brothers and sisters, more comprehensive temperature is ranked the elementary program commending result of acquisition; In conjunction with elementary program commending result and program Relational database, generate program commending result.
In the recommend method of described TV programme, described step D specifically comprises:
D1, program commending results conversion is become to channel bank list, and merge with recommendation of the channels result, generate recommendation of the channels list;
D2, recommendation of the channels list is carried out to repeated program selection, filtration treatment, according to default situation, adjust the sequence of channel;
D3, in conjunction with user channel custom model, recommendation of the channels list is revised, generated final television channel lists.
A system that realizes the recommend method of described TV programme, it comprises client and service end;
Described client is used for obtaining user's diving tower data and is transferred to service end;
Described service end comprises:
Log analysis module, obtains user behavior data for described user's diving tower data are carried out to log analysis;
Recommendation of the channels module, the recommendation of the channels result obtaining for carry out VHI, user preferences and similarity analysis according to user's diving tower data;
Program commending module, obtains program commending result for described user behavior data being carried out to feature weight analysis and EPG program data being carried out to similarity analysis;
Combine recommendation module, adjusts television channel lists for comprehensive described recommendation of the channels result and program commending result.
In the recommend method of described TV programme, described log analysis module comprises:
Data processing module, for according to diving tower data analysis user to program watch frequency and viewing time, generate user behavior data;
Memory module, for storing described user behavior data.
In the recommend method of described TV programme, described recommendation of the channels module comprises:
Conceptual data statistical module, for according to described user's diving tower data statistics watch the channel that number is maximum, generate popular channel bank name;
Statistical module, watches duration the sequence of each channel for adding up unique user in each period at times, generates user preferred channels rank;
Collaborative filtering recommending module, carries out cluster analysis for the behavior of watching of counting user to user, generates similar users table; According to user preferred channels rank correction similar users table, then carry out user collaborative and filter generation collaborative filtering recommendation of the channels table;
Custom MBM, for being accustomed to modeling according to VHI rank and user preferred channels rank, generates user channel custom model;
The first result Fusion Module, for generating recommendation of the channels result after VHI rank, user preferred channels rank and collaborative filtering recommendation of the channels table being merged, filter, sorting.
In the recommend method of described TV programme, described program commending module comprises:
Feature weight analysis module, for carrying out feature extraction and feature weight analysis, generating feature weight database to EPG program data;
User characteristics MBM, for to user behavior data according to time attenuation function, in conjunction with feature weight Database user characteristics model;
Programme content MBM, for extract the content information of program from EPG program data, carries out modeling according to described content information and feature weight database, generates programs feature model;
Program associated recommendation module, for described EPG program data is carried out to correlation calculations, generates program Relational database;
Characteristic model matching module, be used for obtaining real-time hot broadcast channel and sort by temperature, in conjunction with EPG program data, extract hot broadcast program, described programs feature model is mated with user characteristics model and obtain degree of correlation seniority among brothers and sisters, more comprehensive temperature seniority among brothers and sisters obtains elementary program commending result
The second result Fusion Module, for generating program commending result in conjunction with elementary program commending result and program Relational database.
In the recommend method of described TV programme, described combine recommendation module comprises:
Program channel conversion module, for becoming channel bank list by program commending results conversion;
Recommend Fusion Module, for channel bank list and recommendation of the channels result are merged, generate recommendation of the channels list;
Gravity treatment filtering module, for carrying out repeated program selection, filtration treatment to recommendation of the channels list;
Processing module, for adjusting the sequence of channel according to default situation;
Habit modification module, in conjunction with user channel custom model, recommendation of the channels list being revised, generates final television channel lists.
Compared to prior art, the recommend method of TV programme provided by the invention and system thereof, by user's diving tower data analysis is processed, the recommendation of the channels result of acquisition based on user preferences and custom and obtain program commending result based on feature weight and similarity analysis, synthesizer channel recommendation results and program commending result are adjusted television channel lists, in current programme televised live, provide spectators' most likely wanted to see program, the accuracy that improves user's preference degree analysis, has improved programs recommended quality.
Accompanying drawing explanation
Fig. 1 is the flow chart of the recommend method preferred embodiment of TV programme of the present invention.
Fig. 2 is the structured flowchart of the system preferred embodiment of the present invention's recommend method of realizing TV programme.
Embodiment
The invention provides a kind of recommend method and system thereof of TV programme, user's diving tower data analysis is obtained to diving tower behavior, can really reflect the behavior that user initiatively occurs, in conjunction with live temperature, broadcasting program commending, the problem of commending system cold start-up can be solved, and personalized recommendation effect can be improved.Meanwhile, in conjunction with recommending to have produced the brand-new programme televised live way of recommendation based on recommendation of the channels and the personalized interest based on programme content, in current programme televised live, analyze spectators' most likely wanted to see program, and unify to be converted to channel list mode and recommend user.The recommendation of binding channel and programme content reaches complementary effect, can reflect more exactly user's hobby.Also increase in addition particular program processing, considered user's the custom of watching, to improve the accuracy of user behavior analysis, guaranteed programs recommended quality.The present invention changes the feature fast, requirement of real-time is high according to programme televised live, adopts off-line modeling, the methods such as online correlation analysis, and based on watching temperature rank, the order of programs recommended list is adjusted in personalization in real time, the requirement of processing to meet programme televised live.
For making object of the present invention, technical scheme and effect clearer, clear and definite, referring to accompanying drawing, developing simultaneously, the present invention is described in more detail for embodiment.Should be appreciated that specific embodiment described herein, only in order to explain the present invention, is not intended to limit the present invention.
Refer to Fig. 1, the flow chart of its recommend method preferred embodiment that is TV programme of the present invention.The recommend method of the TV programme that as shown in Figure 1, the present embodiment provides comprises the following steps:
S100, client are obtained user's diving tower data and are transferred to service end.
S200, service end are carried out log analysis to described user's diving tower data and are obtained user behavior data, according to user's diving tower data, carry out the recommendation of the channels result that VHI, user preferences and similarity analysis obtain.
S300, service end carry out feature weight analysis and EPG program data are carried out to similarity analysis obtaining program commending result to described behavioral data.
S400, comprehensive described recommendation of the channels result and program commending result are adjusted television channel lists.
In the present embodiment, described client is TV set-top box or TV internal control system, the diving tower behavior of energy recording user, described diving tower behavior comprises that user enters the system time of certain channel (jumping into the time) and the TV station of this channel of entering numbering, and user exits the system time (jumping out the time) of certain channel etc.These behaviors are collected and gathered and be processed into unified form, form user's diving tower data.
In described step S100, client can be compressed packing storage after obtaining user's diving tower data, can by proprietary communication mode, upload user's diving tower data to service end termly afterwards.
Described service end is long distance service system, be used for user's diving tower data and EPG(ElectronicProgramGuide, electric program menu) program data carries out corresponding analyzing and processing, specifically comprises: log analysis flow process, recommendation of the channels flow process, program commending flow process and combine recommendation flow process.Should be understood that, because client is TV set-top box or TV internal control system, the environment for use of user's diving tower data that each client reports based on television set and beholder's type and there is specificity.Service end can be carried out comprehensive statistics analysis to user's diving tower data of each client, then in conjunction with the specificity of each client user's diving tower data, generates corresponding television channel lists and feed back to corresponding client.Service end is before user's diving tower data analysis that client is reported is processed, can first will carry out aggregation process to these user's diving tower data, data requirement is verified, the data of illegal form and error of transmission are rejected in screening, enter afterwards above-mentioned four flow processs again and analyze.These four flow processs are embodied in above-mentioned steps S200 to the concrete scheme of step S400, below in conjunction with step S200, set forth respectively this four flow processs to the concrete scheme of step S400.
One, log analysis flow process
In described step S200, service end is carried out log analysis to diving tower data and is obtained user behavior data and specifically comprise:
Step 201, according to diving tower data analysis user to program watch frequency and viewing time, generate user behavior data.Concrete analysis mode comprises following several:
1, historical behavior correlation analysis---the correlation of program during analysis user historical behavior and diving tower.
2, EPG program objective evaluation---temperature and spectators by EPG program itself evaluate, and after objective rank, as the criteria of quality evaluation of program itself, and set the program that quality is high and liked by spectators.
3, similar users agreement filters---all users' historical behavior is done to the analysis of collaborative filtering, and the preference degree of judgement user to program, thus judgement user watches this type program possibility.
4, user habit analysis---according to the periodicity judgement of the specific custom of user.
5, special behavior is processed---and think and after user starts shooting, select the top line watched, particularly watching the top line of default duration is that user wants to watch.Follow-up selection has watched the program of default duration also can think the program category that user likes.Described default duration is 80% or 90% of this program total time.Program when jumping out, if jumped out when program finishes, thinks that this program watches the program of behavior for user, if user just jumps out in advance not finishing watching this program, user's the behavior of watching does not occur.About not watching the judgement of behavior can also adopt the time of staying, user is less than or equal to 5% of this program total time in the time of staying of this program, thinks behavior does not occur to watch.
The processing that can also be connected to continuous user's diving tower data is processed in above-mentioned special behavior.If user, for two diving tower behaviors of Yi Ge TV station in a Preset Time section, thinks that two are watched behavior continuous, being equivalent to does not have diving tower.For example, while playing TV play, because advertisement user jumps out HNTV, proceeds to Anhui STV, after approximately 2 minutes, advertisement finishes, and user is switched go back to again HNTV, so just thinks and there is no zapping.
Step 202, store described user behavior data.User behavior data provides reference data for follow-up flow process.
Log analysis flow process be analysis user to diving tower during the subjectivity of all programs, judge whether these programs user watched (watching behavior), the program that the program analysis of watching is wanted and liked watching for user.What user's diving tower data pin was right is channel, the user behavior data generating through log analysis flow process for be program.The content of program can inquire from existing EPG program data.User behavior data is the watch behavior of user to program.
Two, recommendation of the channels flow process
In described step S200, the recommendation of the channels result of carrying out VHI, user preferences and similarity analysis acquisition according to user's diving tower data specifically comprises:
The user group of all ages and classes layer likes the program category difference of watching; As old man likes Chinese folk art forms class program or city play more, young man likes idol acute, and child likes cartoon.Based on the study and work time, the user group that different time sections is watched is different; As old man need not go to work, watch program general daytime; How in the afternoon classes are over for child early, and cartoon 5,6 broadcastings; Young man goes to work daytime, After Hours generally watches the TV play of 8 broadcastings evening.Possible certain channel program from morning to night can meet the user's of all ages and classes layer needs, and therefore, VHI is the temperature rank for all users.The program that possible different channel is play at different time meets respectively the hobby of different user colony, and therefore, user preferred channels rank can sum up user and be more prone to watch which channel.May a TV programme can obtain old man and youthful favor, therefore, similar users table can obtain the program category of all liking between different user colony simultaneously.
The hobby that can embody more accurately user through revised similar users table, program category is now all user preferences.For example party A-subscriber likes X, Y type program, and party B-subscriber likes X, Y, Z-type program.X, Y, Z-type program are all in revised similar users table, and Z-type program and X, Y type program are similar, and possible user A does not know Z-type program, watch behavior so not relevant.After filtering by user collaborative, can think that Z-type program also meets party A-subscriber's hobby, can provide advisory opinion to party A-subscriber.User collaborative filters is a kind of recommendation of carrying out based on one group of identical user of interest, and collaborative filtering recommendation of the channels table is to produce the recommendation list to targeted customer according to the user's similar to targeted customer's interest preference information, by the period, distinguishes.
Described recommendation of the channels result is the content that combines VHI rank, user preferred channels rank and collaborative filtering recommendation of the channels table, and by the personalized recommendation result that the period is distinguished, it can reflect the fancy grade of user to different channel more accurately.
The custom modeling of user channel is mainly by individual subscriber is gone out to some mannerisms that user watches this channel to the behavioural analysis of watching of channel, comprises periodically custom and persistence custom.Persistence custom refers to that certain user turns on television set at every turn and will be locked in some channel and watch, and may lose interest in, but there is no zapping to channel at the programme content of broadcasting, and there is no selectivity.If analyze in this case the program channel of user preferences, do not meet on the contrary user's request.Periodically the analysis purpose of custom and persistence custom are similar, by watch the periods rules of historical time Sequence clustering to sum up user's the behavior of watching to user channel.The custom periodic model that the modeling of user channel custom obtains is that a plurality of users for a television set may occur, certain user fixedly watches behavior at the specific channel of specific time period.
Three, program commending flow process
Described step S300 specifically comprises:
Step 301, EPG program data is carried out to feature extraction and feature weight analysis, generating feature weight database; To user behavior data according to time attenuation function, in conjunction with feature weight Database user characteristics model;
In EPG program data content, extracting feature, and according to the information feature of TV programme and TF-IDF(term frequency – inverse document frequency) algorithm comes Feature Weighting, and feature weight database is used in later stage modeling.Described feature refers to for describing the Feature Words of program, as personage, place, language, type etc.If what program was play is film, feature can be performer's name, director's name, shooting ground, swordsman etc.To Feature Weighting, be the number of times that in the different programs of statistics, these features occur, number of times is more, weight is heavier, thereby can analysis user tend to certain feature, as certain performer or director's play works, or tends to the movie and television play of swordsman's type.According to user's behavioral data, extract the program that it is watched, then according to the feature weight in time attenuation function and feature weight database, set up user characteristics model.
Step 302, from EPG program data, extract the content information of program, according to described content information and feature weight database, carry out modeling, generate programs feature model.
Step 303, described EPG program data is carried out to correlation calculations, generate program Relational database.
Described correlation calculations specifically comprises: 1, to the content in EPG program data, according to the similarity of content, the characteristic model of matching content, generates the correlated results of content-based similarity.2,, according to user behavior data, by collaborative filtering similarity, calculate program similarity.3, the correlated results of content-based similarity and the fusion of program similarity are obtained to program Relational database.
Step 304, obtain real-time hot broadcast channel and by temperature sequence, in conjunction with EPG program data, extract hot broadcast program, described programs feature model mate with user characteristics model and obtain degree of correlation seniority among brothers and sisters, more comprehensive temperature is ranked the elementary program commending of acquisition; In conjunction with elementary program commending result and program Relational database, generate program commending result.
When user's Real time request is recommended, obtain online real-time hot broadcast channel, all channels, by temperature sequence, are then extracted to the current hot broadcast program of broadcasting in conjunction with EPG program data.After comprehensive elementary program commending result and program Relational database, filter again, duplicate removal and sequence, finally generate live program commending result
Four, combine recommendation flow process
Combine recommendation is mainly synthesizer channel recommendation results and program commending result.Described step S400 specifically comprises:
Step 401, program commending results conversion is become to channel bank list, and merge with recommendation of the channels result, generate recommendation of the channels list.
What based on program commending result, relate to is program permutation table, what recommendation of the channels result related to is channel information, finally to adjust television channel lists, therefore need in EPG programme, inquire about, find the channel corresponding with program in program commending result, the program permutation table of program commending result is converted to channel bank list.
During by channel bank list and the fusion of recommendation of the channels result, need first to the two, give different weights, by normalization in order, the power of then composing is added, and obtains new recommendation of the channels list after rearrangement.
Step 402, recommendation of the channels list is carried out to repeated program selection, filtration treatment, according to default situation, adjust the sequence of channel.
Because different channel can broadcast identical program, the broadcast progress of each channel is different, and as a TV play, some channels are played to the 10th collection, and some channels have been played to the 20th collection.Can be historical by suitable channel in advance according to watching of user, for example user has watched certain TV play to a certain collection, the channel of playing this next collection of TV play is put into recommendation of the channels list front end.Other channel banks of playing identical TV play are at recommendation of the channels list tail end.
The sequence of adjusting channel according to default situation is the progress for each current broadcast items of TV station, and the channel of the program closing to an end is filled into recommendation of the channels list tail end.After this program finishes, can show the program that user likes in addition at once, can make user proceed to as early as possible in other programs, while avoiding user to select, lose time.
Step 403, in conjunction with user channel custom model, recommendation of the channels list is revised, generated final television channel lists.
User channel custom model is the statistics of watching custom done special to user, and these watch custom to have stationarity or timeliness, as certain user watches TV, just fix one or several TV station, and select insensitive to program itself; Fixedly watch Yi Ge TV station again or in some period, as 7 news hookups of watching CCTV1 in evening; Or user likes the serial of watching certain TV station to play.In this case, in order to respect user's the custom of watching, according to user's custom, recommendation of the channels list is adjusted, finally generate television channel lists.
What suppose that recommendation of the channels list foremost lists is the channel for TV plays and movies that user often sees recently, but system time is shown as point in evenings 7, according to user channel custom model, knows that user is now more prone to watch the news hookup of CCTV1; Therefore, with user channel custom model, be as the criterion, the television channel lists foremost of generation is CCTV1, is then this channel for TV plays and movies.
The embodiment of the present invention is also corresponding provides a kind of for realizing the system of the recommend method of described TV programme, and as shown in Figure 2, it comprises client 10 and service end 20.
Described client 10 is for obtaining user's diving tower data and being transferred to service end 20, and it is television set top box or inside television control system.Described client 10 comprises acquisition module 110 and reporting module 120, and described acquisition module, for the diving tower behavior of recording user, is collected and gathered formation diving tower data diving tower behavior, and compresses packing storage.Reporting module 120 is uploaded user's diving tower data to service end.
Described service end 20 comprises log analysis module 210, recommendation of the channels module 220, program commending module 230 and combine recommendation module 240, described log analysis module 210, recommendation of the channels module 220, combine recommendation module 240, program commending module 230 connect successively, and described log analysis module 210 connects program recommending module 230.
Described log analysis module 210 obtains for described user's diving tower data being carried out to log analysis the user behavior data whether user watches TV programme.Recommendation of the channels module 220 carries out according to user's diving tower data the recommendation of the channels result that VHI, user preferences and similarity analysis obtain.230 pairs of described user behavior datas of program commending module carry out feature weight analysis and EPG program data are carried out to similarity analysis obtaining program commending result.The comprehensive described recommendation of the channels result of combine recommendation module 240 and program commending result are adjusted television channel lists.
Wherein, described log analysis module 210 comprises:
Data processing module, for according to diving tower data analysis user to program watch frequency and viewing time, generate user behavior data;
Memory module, for storing described user behavior data.
Described recommendation of the channels module 220 comprises:
Conceptual data statistical module, for according to described user's diving tower data statistics watch the channel that number is maximum, generate popular channel bank name;
Statistical module, watches duration the sequence of each channel for adding up unique user in each period at times, generates user preferred channels rank;
Collaborative filtering recommending module, carries out cluster analysis for the behavior of watching of counting user to user, generates similar users table; According to user preferred channels rank correction similar users table, then carry out user collaborative and filter generation collaborative filtering recommendation of the channels table;
Custom MBM, for being accustomed to modeling according to VHI rank and user preferred channels rank, generates user channel custom model;
The first result Fusion Module, for generating recommendation of the channels result after VHI rank, user preferred channels rank and collaborative filtering recommendation of the channels table being merged, filter, sorting.
Described program commending module 230 comprises:
Feature weight analysis module, for carrying out feature extraction and feature weight analysis, generating feature weight database to EPG program data;
User characteristics MBM, for to user behavior data according to time attenuation function, in conjunction with feature weight Database user characteristics model;
Programme content MBM, for extract the content information of program from EPG program data, carries out modeling according to described content information and feature weight database, generates programs feature model;
Program associated recommendation module, for described EPG program data is carried out to correlation calculations, generates program Relational database;
Characteristic model matching module, be used for obtaining real-time hot broadcast channel and sort by temperature, in conjunction with EPG program data, extract hot broadcast program, described programs feature model is mated with user characteristics model and obtain degree of correlation seniority among brothers and sisters, more comprehensive temperature seniority among brothers and sisters obtains elementary program commending result
The second result Fusion Module, for generating program commending result in conjunction with elementary program commending result and program Relational database.
Described combine recommendation module 240 comprises:
Program channel conversion module, for becoming channel bank list by program commending results conversion;
Recommend Fusion Module, for channel bank list and recommendation of the channels result are merged, generate recommendation of the channels list;
Gravity treatment filtering module, for carrying out repeated program selection, filtration treatment to recommendation of the channels list;
Processing module, for adjusting the sequence of channel according to default situation;
Habit modification module, in conjunction with user channel custom model, recommendation of the channels list being revised, generates final television channel lists.
In sum, the recommend method of TV programme of the present invention and system thereof, propose that a kind of synthesizer channel is recommended and the combine recommendation mode of program commending, user's diving tower data analysis is processed to obtain user initiatively occurs watches behavior, the hobby of confirmation user's program; In conjunction with popular program, user preferences and user's similarity, programme televised live is analyzed and obtained recommendation of the channels result, user behavior data is carried out feature weight analysis and EPG program data is carried out to similarity analysis obtaining program commending result, comprehensive recommendation of the channels result and the personalized program commending result based on program based on channel, in current programme televised live, provide spectators' most likely wanted to see program, be converted to television channel lists and recommend user, this way of recommendation has improved the accuracy of user's preference degree analysis, has improved programs recommended quality.
Be understandable that, for those of ordinary skills, can be equal to replacement or change according to technical scheme of the present invention and inventive concept thereof, and all these changes or replacement all should belong to the protection range of the appended claim of the present invention.
Claims (10)
1. a recommend method for TV programme, is characterized in that, comprising:
A, client are obtained user's diving tower data and are transferred to service end;
B, service end are carried out log analysis to described user's diving tower data and are obtained user behavior data, according to user's diving tower data, carry out the recommendation of the channels result that VHI, user preferences and similarity analysis obtain;
C, service end carry out feature weight analysis and EPG program data are carried out to similarity analysis obtaining program commending result to described user behavior data;
D, comprehensive described recommendation of the channels result and program commending result are adjusted television channel lists.
2. the recommend method of TV programme according to claim 1, is characterized in that, in described step B, service end is carried out log analysis to diving tower data and obtained user behavior data and specifically comprise:
B1, according to diving tower data analysis user to program watch frequency and viewing time, generate user behavior data;
B2, store described user behavior data.
3. the recommend method of TV programme according to claim 2, is characterized in that, in described step B, the recommendation of the channels result of carrying out VHI, user preferences and similarity analysis acquisition according to user's diving tower data specifically comprises:
B21, according to described user's diving tower, the channel that number is maximum is watched in data statistics, generates popular channel bank name; Statistics unique user is watched duration the sequence of each channel in each period, generate user preferred channels rank; The behavior of watching of counting user is carried out cluster analysis to user, generates similar users table;
B22, according to user preferred channels rank correction similar users table, then carry out user collaborative and filter to generate collaborative filtering recommendation of the channels table;
B23, after being merged, filter, sorted, VHI rank, user preferred channels rank and collaborative filtering recommendation of the channels table generate recommendation of the channels result; According to VHI rank and user preferred channels rank, be accustomed to modeling, generate user channel custom model.
4. the recommend method of TV programme according to claim 1, is characterized in that, described step C specifically comprises:
C1, EPG program data is carried out to feature extraction and feature weight analysis, generating feature weight database; To user behavior data according to time attenuation function, in conjunction with feature weight Database user characteristics model;
C2, from EPG program data, extract the content information of program, according to described content information and feature weight database, carry out modeling, generate programs feature model;
C3, described EPG program data is carried out to correlation calculations, generate program Relational database;
C4, obtain real-time hot broadcast channel and by temperature sequence, in conjunction with EPG program data, extract hot broadcast program, described programs feature model mate with user characteristics model and obtain degree of correlation seniority among brothers and sisters, more comprehensive temperature is ranked the elementary program commending result of acquisition; In conjunction with elementary program commending result and program Relational database, generate program commending result.
5. the recommend method of TV programme according to claim 4, is characterized in that, described step D specifically comprises:
D1, program commending results conversion is become to channel bank list, and merge with recommendation of the channels result, generate recommendation of the channels list;
D2, recommendation of the channels list is carried out to repeated program selection, filtration treatment, according to default situation, adjust the sequence of channel;
D3, in conjunction with user channel custom model, recommendation of the channels list is revised, generated final television channel lists.
6. for realizing a system for the recommend method of TV programme claimed in claim 1, it is characterized in that, comprise client and service end;
Described client is used for obtaining user's diving tower data and is transferred to service end;
Described service end comprises:
Log analysis module, obtains user behavior data for described user's diving tower data are carried out to log analysis;
Recommendation of the channels module, the recommendation of the channels result obtaining for carry out VHI, user preferences and similarity analysis according to user's diving tower data;
Program commending module, obtains program commending result for described user behavior data being carried out to feature weight analysis and EPG program data being carried out to similarity analysis;
Combine recommendation module, adjusts television channel lists for comprehensive described recommendation of the channels result and program commending result.
7. the recommend method of TV programme according to claim 6, is characterized in that, described log analysis module comprises:
Data processing module, for according to diving tower data analysis user to program watch frequency and viewing time, generate user behavior data;
Memory module, for storing described user behavior data.
8. the recommend method of TV programme according to claim 7, is characterized in that, described recommendation of the channels module comprises:
Conceptual data statistical module, for according to described user's diving tower data statistics watch the channel that number is maximum, generate popular channel bank name;
Statistical module, watches duration the sequence of each channel for adding up unique user in each period at times, generates user preferred channels rank;
Collaborative filtering recommending module, carries out cluster analysis for the behavior of watching of counting user to user, generates similar users table; According to user preferred channels rank correction similar users table, then carry out user collaborative and filter generation collaborative filtering recommendation of the channels table;
Custom MBM, for being accustomed to modeling according to VHI rank and user preferred channels rank, generates user channel custom model;
The first result Fusion Module, for generating recommendation of the channels result after VHI rank, user preferred channels rank and collaborative filtering recommendation of the channels table being merged, filter, sorting.
9. the recommend method of TV programme according to claim 6, is characterized in that, described program commending module comprises:
Feature weight analysis module, for carrying out feature extraction and feature weight analysis, generating feature weight database to EPG program data;
User characteristics MBM, for to user behavior data according to time attenuation function, in conjunction with feature weight Database user characteristics model;
Programme content MBM, for extract the content information of program from EPG program data, carries out modeling according to described content information and feature weight database, generates programs feature model;
Program associated recommendation module, for described EPG program data is carried out to correlation calculations, generates program Relational database;
Characteristic model matching module, be used for obtaining real-time hot broadcast channel and sort by temperature, in conjunction with EPG program data, extract hot broadcast program, described programs feature model is mated with user characteristics model and obtain degree of correlation seniority among brothers and sisters, more comprehensive temperature seniority among brothers and sisters obtains elementary program commending result;
The second result Fusion Module, for generating program commending result in conjunction with elementary program commending result and program Relational database.
10. the recommend method of TV programme according to claim 6, is characterized in that, described combine recommendation module comprises:
Program channel conversion module, for becoming channel bank list by program commending results conversion;
Recommend Fusion Module, for channel bank list and recommendation of the channels result are merged, generate recommendation of the channels list;
Gravity treatment filtering module, for carrying out repeated program selection, filtration treatment to recommendation of the channels list;
Processing module, for adjusting the sequence of channel according to default situation;
Habit modification module, in conjunction with user channel custom model, recommendation of the channels list being revised, generates final television channel lists.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310355368.5A CN103546773B (en) | 2013-08-15 | 2013-08-15 | The recommendation method and its system of TV programme |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310355368.5A CN103546773B (en) | 2013-08-15 | 2013-08-15 | The recommendation method and its system of TV programme |
Publications (2)
Publication Number | Publication Date |
---|---|
CN103546773A true CN103546773A (en) | 2014-01-29 |
CN103546773B CN103546773B (en) | 2017-07-11 |
Family
ID=49969764
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201310355368.5A Expired - Fee Related CN103546773B (en) | 2013-08-15 | 2013-08-15 | The recommendation method and its system of TV programme |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103546773B (en) |
Cited By (45)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104038790A (en) * | 2014-06-30 | 2014-09-10 | 北京奇艺世纪科技有限公司 | Video playing list updating method and system |
CN104053023A (en) * | 2014-06-13 | 2014-09-17 | 海信集团有限公司 | Method and device for determining video similarity |
CN104065981A (en) * | 2014-06-20 | 2014-09-24 | 海信集团有限公司 | Method and device for recommending videos |
CN104079997A (en) * | 2014-07-10 | 2014-10-01 | 东莞中山大学研究院 | Recommending system for digital television personalized programs and method thereof |
CN104394471A (en) * | 2014-11-19 | 2015-03-04 | 四川长虹电器股份有限公司 | Method for intelligently recommending favorite program to user |
CN104539981A (en) * | 2014-11-26 | 2015-04-22 | 四川长虹电器股份有限公司 | Real-time hot television channel recommending system and method |
CN104639984A (en) * | 2015-03-12 | 2015-05-20 | 北京京东方多媒体科技有限公司 | Network set top box and working method thereof |
CN104766607A (en) * | 2015-03-05 | 2015-07-08 | 广州视源电子科技股份有限公司 | Television program recommendation method and system |
CN104796734A (en) * | 2015-03-20 | 2015-07-22 | 四川长虹电器股份有限公司 | Real-time interactive smart television program combined recommendation system and method |
CN104811758A (en) * | 2015-03-30 | 2015-07-29 | 腾讯科技(北京)有限公司 | Program list providing method and device |
CN104951956A (en) * | 2014-03-31 | 2015-09-30 | Tcl集团股份有限公司 | Film ranking method and device |
CN104954821A (en) * | 2015-06-24 | 2015-09-30 | 北京酷云互动科技有限公司 | Program correlation calculating method and system |
CN104994444A (en) * | 2015-07-06 | 2015-10-21 | 无锡天脉聚源传媒科技有限公司 | Method and device for displaying video summary information |
CN105069049A (en) * | 2015-07-23 | 2015-11-18 | 四川长虹电器股份有限公司 | Method for screening and correlation arrangement of related programs in video-on-demand library |
CN105096152A (en) * | 2014-05-20 | 2015-11-25 | 阿里巴巴集团控股有限公司 | Commodity popularity-based operation execution method and device |
CN105321442A (en) * | 2014-06-23 | 2016-02-10 | 卡西欧计算机株式会社 | Information evaluation apparatus and information evaluation method |
CN105376649A (en) * | 2015-11-24 | 2016-03-02 | 江苏有线技术研究院有限公司 | Set top box blind operation method and system for realizing accurate combination recommendation |
CN105430504A (en) * | 2015-11-27 | 2016-03-23 | 中国科学院深圳先进技术研究院 | Family member mix identification method and system based on television watching log mining |
CN105430444A (en) * | 2015-11-17 | 2016-03-23 | 深圳创维数字技术有限公司 | Program recommendation method and digital TV receiving terminal |
CN105701141A (en) * | 2015-11-26 | 2016-06-22 | 中国传媒大学 | Query manner of image material library |
CN105812845A (en) * | 2014-12-30 | 2016-07-27 | Tcl集团股份有限公司 | Media resource pushing method, media resource pushing system, and media player based on Android system |
CN105898423A (en) * | 2015-12-08 | 2016-08-24 | 乐视网信息技术(北京)股份有限公司 | Video pushing method and system and server |
CN106294564A (en) * | 2016-07-27 | 2017-01-04 | 乐视控股(北京)有限公司 | A kind of video recommendation method and device |
CN106454400A (en) * | 2016-09-20 | 2017-02-22 | 广东工业大学 | Time-sequenced theme preference prediction method oriented to network television user |
CN107623866A (en) * | 2017-11-06 | 2018-01-23 | 深圳Tcl数字技术有限公司 | Video recommendation method, device and the storage medium of television set |
CN108093304A (en) * | 2017-12-29 | 2018-05-29 | 东莞市纽格力信息技术有限公司 | A kind of intelligent recommendation method and system based on user's custom |
CN108307208A (en) * | 2018-01-10 | 2018-07-20 | 武汉斗鱼网络科技有限公司 | Calculate method, storage medium, equipment and the system of direct broadcasting room similarity |
CN108304888A (en) * | 2018-02-28 | 2018-07-20 | 广州虎牙信息科技有限公司 | Platform user sorting technique and computer storage media, terminal is broadcast live |
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 |
CN108924655A (en) * | 2018-08-01 | 2018-11-30 | 北京奇艺世纪科技有限公司 | A kind of videoconference client navigation generation method and device |
CN109033233A (en) * | 2018-06-29 | 2018-12-18 | 武汉斗鱼网络科技有限公司 | A kind of direct broadcasting room recommended method, storage medium, electronic equipment and system |
CN109522062A (en) * | 2018-11-16 | 2019-03-26 | 北京微播视界科技有限公司 | Display data determines method, applications client and electronic equipment |
CN109618219A (en) * | 2018-12-28 | 2019-04-12 | 中国移动通信集团江苏有限公司 | Fast channel switching method, device, equipment and the storage medium of Web TV |
CN109688458A (en) * | 2019-01-14 | 2019-04-26 | 四川长虹电器股份有限公司 | The implementation method of smart television cloud desktop operation system based on big data algorithm |
CN109740048A (en) * | 2018-12-12 | 2019-05-10 | 中国联合网络通信集团有限公司 | A kind of course recommended method and device |
CN110191363A (en) * | 2019-05-31 | 2019-08-30 | 电子科技大学 | A kind of recommended models towards family group user |
CN110392307A (en) * | 2018-04-17 | 2019-10-29 | 富泰华工业(深圳)有限公司 | TV programme filtration system and method |
CN110769286A (en) * | 2019-11-06 | 2020-02-07 | 山东科技大学 | Channel-based recommendation method and device and storage medium |
CN111209470A (en) * | 2018-11-21 | 2020-05-29 | 国家新闻出版广电总局广播科学研究院 | Personalized content recommendation method and device and storage medium |
CN111836099A (en) * | 2019-04-17 | 2020-10-27 | 百度在线网络技术(北京)有限公司 | Channel switching processing method and sound box |
CN111935513A (en) * | 2020-07-14 | 2020-11-13 | 广东工业大学 | Home user-oriented network television program recommendation method and device |
CN112040318A (en) * | 2020-08-21 | 2020-12-04 | 广州欢网科技有限责任公司 | Personalized program recommendation method and device, storage medium and electronic equipment |
CN112333477A (en) * | 2020-06-02 | 2021-02-05 | 深圳Tcl新技术有限公司 | Program recommendation method, device and equipment and computer storage medium |
CN113505291A (en) * | 2021-05-27 | 2021-10-15 | 成都数博视科技有限公司 | Intelligent content recommendation system based on user behavior data |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1756336A (en) * | 2004-09-29 | 2006-04-05 | 松下电器产业株式会社 | Television channel commending system and commending method |
US20070174862A1 (en) * | 2005-12-26 | 2007-07-26 | Mitsubishi Electric Corporation | Recommended program search device and recommended program search method |
CN101163196A (en) * | 2007-11-20 | 2008-04-16 | 腾讯科技(深圳)有限公司 | Method, system and equipment to recommend program to television users |
CN102123308A (en) * | 2010-01-06 | 2011-07-13 | 索尼公司 | Information processing apparatus, information processing method, and program |
CN102769781A (en) * | 2012-07-17 | 2012-11-07 | 青岛海信传媒网络技术有限公司 | Method and device for recommending television program |
-
2013
- 2013-08-15 CN CN201310355368.5A patent/CN103546773B/en not_active Expired - Fee Related
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1756336A (en) * | 2004-09-29 | 2006-04-05 | 松下电器产业株式会社 | Television channel commending system and commending method |
US20070174862A1 (en) * | 2005-12-26 | 2007-07-26 | Mitsubishi Electric Corporation | Recommended program search device and recommended program search method |
CN101163196A (en) * | 2007-11-20 | 2008-04-16 | 腾讯科技(深圳)有限公司 | Method, system and equipment to recommend program to television users |
CN102123308A (en) * | 2010-01-06 | 2011-07-13 | 索尼公司 | Information processing apparatus, information processing method, and program |
CN102769781A (en) * | 2012-07-17 | 2012-11-07 | 青岛海信传媒网络技术有限公司 | Method and device for recommending television program |
Cited By (67)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104951956A (en) * | 2014-03-31 | 2015-09-30 | Tcl集团股份有限公司 | Film ranking method and device |
CN105096152A (en) * | 2014-05-20 | 2015-11-25 | 阿里巴巴集团控股有限公司 | Commodity popularity-based operation execution method and device |
CN104053023B (en) * | 2014-06-13 | 2018-11-30 | 海信集团有限公司 | A kind of method and device of determining video similarity |
CN104053023A (en) * | 2014-06-13 | 2014-09-17 | 海信集团有限公司 | Method and device for determining video similarity |
CN104065981A (en) * | 2014-06-20 | 2014-09-24 | 海信集团有限公司 | Method and device for recommending videos |
CN105321442A (en) * | 2014-06-23 | 2016-02-10 | 卡西欧计算机株式会社 | Information evaluation apparatus and information evaluation method |
CN104038790A (en) * | 2014-06-30 | 2014-09-10 | 北京奇艺世纪科技有限公司 | Video playing list updating method and system |
CN104038790B (en) * | 2014-06-30 | 2018-04-13 | 北京奇艺世纪科技有限公司 | Video play lists update method and system |
CN104079997A (en) * | 2014-07-10 | 2014-10-01 | 东莞中山大学研究院 | Recommending system for digital television personalized programs and method thereof |
CN104394471A (en) * | 2014-11-19 | 2015-03-04 | 四川长虹电器股份有限公司 | Method for intelligently recommending favorite program to user |
CN104539981A (en) * | 2014-11-26 | 2015-04-22 | 四川长虹电器股份有限公司 | Real-time hot television channel recommending system and method |
CN105812845A (en) * | 2014-12-30 | 2016-07-27 | Tcl集团股份有限公司 | Media resource pushing method, media resource pushing system, and media player based on Android system |
CN105812845B (en) * | 2014-12-30 | 2019-09-17 | Tcl集团股份有限公司 | A kind of media resource method for pushing, system and the media player based on android system |
CN104766607A (en) * | 2015-03-05 | 2015-07-08 | 广州视源电子科技股份有限公司 | Television program recommendation method and system |
CN104639984A (en) * | 2015-03-12 | 2015-05-20 | 北京京东方多媒体科技有限公司 | Network set top box and working method thereof |
US9930412B2 (en) | 2015-03-12 | 2018-03-27 | Boe Technology Group Co., Ltd. | Network set-top box and its operating method |
CN104796734A (en) * | 2015-03-20 | 2015-07-22 | 四川长虹电器股份有限公司 | Real-time interactive smart television program combined recommendation system and method |
CN104796734B (en) * | 2015-03-20 | 2018-05-01 | 四川长虹电器股份有限公司 | The smart television program combination commending system and method for real-time, interactive |
CN104811758A (en) * | 2015-03-30 | 2015-07-29 | 腾讯科技(北京)有限公司 | Program list providing method and device |
CN104811758B (en) * | 2015-03-30 | 2018-09-04 | 腾讯科技(北京)有限公司 | Programme providing method and device |
CN104954821B (en) * | 2015-06-24 | 2018-06-19 | 北京酷云互动科技有限公司 | A kind of computational methods and its computing system of programming association degree |
CN104954821A (en) * | 2015-06-24 | 2015-09-30 | 北京酷云互动科技有限公司 | Program correlation calculating method and system |
CN104994444A (en) * | 2015-07-06 | 2015-10-21 | 无锡天脉聚源传媒科技有限公司 | Method and device for displaying video summary information |
CN104994444B (en) * | 2015-07-06 | 2018-09-25 | 无锡天脉聚源传媒科技有限公司 | A kind of display methods and device of video summary information |
CN105069049A (en) * | 2015-07-23 | 2015-11-18 | 四川长虹电器股份有限公司 | Method for screening and correlation arrangement of related programs in video-on-demand library |
CN105430444A (en) * | 2015-11-17 | 2016-03-23 | 深圳创维数字技术有限公司 | Program recommendation method and digital TV receiving terminal |
CN105376649A (en) * | 2015-11-24 | 2016-03-02 | 江苏有线技术研究院有限公司 | Set top box blind operation method and system for realizing accurate combination recommendation |
CN105376649B (en) * | 2015-11-24 | 2018-09-14 | 江苏有线技术研究院有限公司 | Realize the blind operating method of the set-top box of accurate combined recommendation and system |
CN105701141A (en) * | 2015-11-26 | 2016-06-22 | 中国传媒大学 | Query manner of image material library |
CN105430504B (en) * | 2015-11-27 | 2019-04-02 | 中国科学院深圳先进技术研究院 | Kinsfolk's structural recognition method and system based on television-viewing Web log mining |
CN105430504A (en) * | 2015-11-27 | 2016-03-23 | 中国科学院深圳先进技术研究院 | Family member mix identification method and system based on television watching log mining |
CN105898423A (en) * | 2015-12-08 | 2016-08-24 | 乐视网信息技术(北京)股份有限公司 | Video pushing method and system and server |
CN106294564A (en) * | 2016-07-27 | 2017-01-04 | 乐视控股(北京)有限公司 | A kind of video recommendation method and device |
CN106454400B (en) * | 2016-09-20 | 2019-09-13 | 广东工业大学 | A kind of timing subject matter preferences prediction technique of network-oriented TV user |
CN106454400A (en) * | 2016-09-20 | 2017-02-22 | 广东工业大学 | Time-sequenced theme preference prediction method oriented to network television user |
CN107623866B (en) * | 2017-11-06 | 2021-04-27 | 深圳Tcl数字技术有限公司 | Video recommendation method and device for television and storage medium |
CN107623866A (en) * | 2017-11-06 | 2018-01-23 | 深圳Tcl数字技术有限公司 | Video recommendation method, device and the storage medium of television set |
CN108093304A (en) * | 2017-12-29 | 2018-05-29 | 东莞市纽格力信息技术有限公司 | A kind of intelligent recommendation method and system based on user's custom |
WO2019136851A1 (en) * | 2018-01-10 | 2019-07-18 | 武汉斗鱼网络科技有限公司 | Method for calculating similarity between live broadcast rooms, storage medium, device, and system |
CN108307208A (en) * | 2018-01-10 | 2018-07-20 | 武汉斗鱼网络科技有限公司 | Calculate method, storage medium, equipment and the system of direct broadcasting room similarity |
CN108307208B (en) * | 2018-01-10 | 2020-02-07 | 武汉斗鱼网络科技有限公司 | Method, storage medium, device and system for calculating similarity between live broadcasts |
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 |
CN108304888A (en) * | 2018-02-28 | 2018-07-20 | 广州虎牙信息科技有限公司 | Platform user sorting technique and computer storage media, terminal is broadcast live |
CN108304888B (en) * | 2018-02-28 | 2020-07-07 | 广州虎牙信息科技有限公司 | Live broadcast platform user classification method, computer storage medium and terminal |
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 |
CN110392307B (en) * | 2018-04-17 | 2021-11-09 | 富泰华工业(深圳)有限公司 | Television program filtering system and method |
CN110392307A (en) * | 2018-04-17 | 2019-10-29 | 富泰华工业(深圳)有限公司 | TV programme filtration system and method |
CN109033233A (en) * | 2018-06-29 | 2018-12-18 | 武汉斗鱼网络科技有限公司 | A kind of direct broadcasting room recommended method, storage medium, electronic equipment and system |
CN108924655A (en) * | 2018-08-01 | 2018-11-30 | 北京奇艺世纪科技有限公司 | A kind of videoconference client navigation generation method and device |
CN108924655B (en) * | 2018-08-01 | 2020-11-17 | 北京奇艺世纪科技有限公司 | Video client navigation generation method and device |
CN109522062A (en) * | 2018-11-16 | 2019-03-26 | 北京微播视界科技有限公司 | Display data determines method, applications client and electronic equipment |
CN111209470A (en) * | 2018-11-21 | 2020-05-29 | 国家新闻出版广电总局广播科学研究院 | Personalized content recommendation method and device and storage medium |
CN111209470B (en) * | 2018-11-21 | 2023-08-04 | 国家新闻出版广电总局广播科学研究院 | Personalized content recommendation method, device and storage medium |
CN109740048A (en) * | 2018-12-12 | 2019-05-10 | 中国联合网络通信集团有限公司 | A kind of course recommended method and device |
CN109618219A (en) * | 2018-12-28 | 2019-04-12 | 中国移动通信集团江苏有限公司 | Fast channel switching method, device, equipment and the storage medium of Web TV |
CN109618219B (en) * | 2018-12-28 | 2021-11-23 | 中国移动通信集团江苏有限公司 | Method, device, equipment and storage medium for quickly changing channel of network television |
CN109688458A (en) * | 2019-01-14 | 2019-04-26 | 四川长虹电器股份有限公司 | The implementation method of smart television cloud desktop operation system based on big data algorithm |
CN111836099A (en) * | 2019-04-17 | 2020-10-27 | 百度在线网络技术(北京)有限公司 | Channel switching processing method and sound box |
CN110191363A (en) * | 2019-05-31 | 2019-08-30 | 电子科技大学 | A kind of recommended models towards family group user |
CN110191363B (en) * | 2019-05-31 | 2020-05-01 | 电子科技大学 | Family group user oriented recommendation method |
CN110769286A (en) * | 2019-11-06 | 2020-02-07 | 山东科技大学 | Channel-based recommendation method and device and storage medium |
CN110769286B (en) * | 2019-11-06 | 2021-04-27 | 山东科技大学 | Channel-based recommendation method and device and storage medium |
CN112333477A (en) * | 2020-06-02 | 2021-02-05 | 深圳Tcl新技术有限公司 | Program recommendation method, device and equipment and computer storage medium |
CN112333477B (en) * | 2020-06-02 | 2022-11-15 | 深圳Tcl新技术有限公司 | Program recommendation method, device and equipment and computer storage medium |
CN111935513A (en) * | 2020-07-14 | 2020-11-13 | 广东工业大学 | Home user-oriented network television program recommendation method and device |
CN112040318A (en) * | 2020-08-21 | 2020-12-04 | 广州欢网科技有限责任公司 | Personalized program recommendation method and device, storage medium and electronic equipment |
CN113505291A (en) * | 2021-05-27 | 2021-10-15 | 成都数博视科技有限公司 | Intelligent content recommendation system based on user behavior data |
Also Published As
Publication number | Publication date |
---|---|
CN103546773B (en) | 2017-07-11 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103546773A (en) | Television program recommendation method and system | |
US20230252048A1 (en) | Feedback loop content recommendation | |
CN104935968B (en) | The method and system of Web TV recommendation program | |
AU2016202425B2 (en) | Automatically recommending content | |
CN102263999B (en) | Face-recognition-based method and system for automatically classifying television programs | |
US20190045273A1 (en) | Enhanced program guide | |
CN103124377B (en) | System for minimizing interactions with at least an input mechanism | |
US20130311408A1 (en) | Determining and Predicting Popularity of Content | |
CN103563275B (en) | Determine the method and apparatus of the use of media resource | |
US20140259038A1 (en) | Demographic prediction channel | |
CN110430471A (en) | It is a kind of based on the television recommendations method and system instantaneously calculated | |
US20030131355A1 (en) | Program guide system | |
CN102780920A (en) | Television program recommending method and system | |
CN102184235A (en) | Set top box-based digital television program recommending method and system | |
CN102630052A (en) | Real time streaming-oriented television program recommendation system | |
CN103686382A (en) | Program recommendation method | |
CN106791964B (en) | Broadcast TV program recommender system and method | |
CN108419134B (en) | Channel recommendation method based on fusion of individual history and group current behaviors | |
CN104219576A (en) | Smart television based play program recommendation method | |
CN111062527B (en) | Video traffic collection prediction method and device | |
CN105376649B (en) | Realize the blind operating method of the set-top box of accurate combined recommendation and system | |
CN106454431A (en) | Method and system for recommending television programs | |
CN106792210A (en) | The sorting technique and system of a kind of TV user | |
CN104219577A (en) | Hybrid real-time television program and network video recommending method based on intelligent televisions | |
US9118965B1 (en) | Virtual video channels |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20170711 |