Broadcasting and TV program recommendation method based on knowledge graph and user microcosmic behaviors

A knowledge map and program recommendation technology, applied in the field of radio and television program recommendation based on knowledge map and user micro-behavior, can solve the problems of ignoring the internal connection of content attributes and not taking into account the differences in user interests, etc., to achieve strong correlation and enhance individuality The effect of chemicalization and precise algorithm

Active Publication Date: 2021-04-30
UNIV OF ELECTRONIC SCI & TECH OF CHINA
View PDF3 Cites 12 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0006] The present invention ignores various behavior types of users in the prior art, does not take into account the user interest differences implied by different feedback behaviors of users, and when mining the dynamic changes of user preferences, the embedding of items obtained from the feature point of view often ignores the In order to solve problems such as the internal connection of content attributes betwe

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Broadcasting and TV program recommendation method based on knowledge graph and user microcosmic behaviors
  • Broadcasting and TV program recommendation method based on knowledge graph and user microcosmic behaviors
  • Broadcasting and TV program recommendation method based on knowledge graph and user microcosmic behaviors

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0107] The present invention proposes a method for recommending radio and television programs based on knowledge graphs and user microbehaviors, such as figure 1 As shown, it specifically includes the following steps:

[0108] Step S1: Build a knowledge graph in the field of radio and television;

[0109] Step S2: Sort out user micro-behavior data: divide user behavior interaction data into continuous micro-behavior and discrete micro-behavior according to behavior duration;

[0110] Step S3: Extract the attribute subgraph of the broadcasting field knowledge map constructed in step S1, and use the random walk method to extract a random walk program sequence Q according to the attribute subgraph; extract a random walk from one attribute subgraph Sequence Q, the random walk sequence Q of all attribute subgraphs together form an item sequence set H k ;

[0111] Step S4: Utilize the user's behavior interaction data sorted out in step S2 to construct a time session-behavior type...

Embodiment 2

[0118] This embodiment is based on the above-mentioned embodiment 1, in order to better realize the present invention, further, as figure 2 As shown, the specific operation is:

[0119]Step S1: Build a knowledge map in the field of radio and television: Crawl the network resources of radio and television programs, form structured data through entity alignment, and use the ontology modeling tool protégé to complete the ontology construction; after the ontology construction is completed, use d2rq to store the original in the relation The data in the large-scale database is converted into the corresponding rdf format, and then stored in the form of a graph database, and then the construction of the knowledge map in the field of broadcasting and television is completed.

[0120] In order to better realize the present invention, further, when constructing the knowledge map in the field of broadcasting and television, the knowledge map is constructed in a top-down manner, and the d...

Embodiment 3

[0123] In this embodiment, on the basis of any one of the above-mentioned embodiments 1-2, in order to better realize the present invention, further, the specific operation of the step S2 is: divide the user's behavior interaction data into continuous Continuous micro-behaviors and discrete micro-behaviors; the continuous micro-behaviors are user behaviors that can last for a certain period of time, including live viewing, on-demand viewing, and search and viewing behaviors; the discrete micro-behaviors are user behaviors that only occur at a certain moment Behaviors, including purchases, favorites, and likes; collect the data detected by the background of the radio and television system to form structured data that records user numbers, media asset numbers, behavior types, behavior timestamps, and behavior durations. The behavior duration of the behavior is the corresponding valid value, and the behavior duration of the discrete behavior type is null.

[0124] Other parts of ...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

PUM

No PUM Login to view more

Abstract

The invention provides a broadcasting and TV program recommendation method based on a knowledge graph and a user microcosmic behavior. The method comprises the steps: combining intem2vec with random walk, training a random walk sequence of an attribute subgraph and a user behavior sequence, and obtaining an embedded vector which integrates the similarity of a program content attribute layer and a user interaction session layer. Then, under the condition that program embedding and classified fusion microscopic behavior embedding are obtained, behavior embedding and corresponding program embedding are spliced according to historical interaction records of a user and a program, semantic representation of behavior programs with the same dimension is obtained through semantic space network mapping, and an embedding sequence of historical behaviors of the user is formed; and finally, based on a Transformer encoding and decoding mechanism, carrying out self-attention encoding and mapping on the historical behavior sequence of the user to obtain user semantic features implicit with dynamic preferences, and decoding the user semantic features by utilizing target program attention to carry out mapping to obtain user semantic preferences.

Description

technical field [0001] The invention belongs to the technical field of radio and television program recommendation, and in particular relates to a radio and television program recommendation method based on knowledge graphs and user microscopic behaviors. Background technique [0002] With the integration of the three-network service of the telecommunication network, the radio and television network and the computer communication network, the services provided by the radio and television network are becoming more and more abundant and updated faster and faster. Due to the increase in the number of TV channels and the emergence of IPTV services and new media services, users exposed to TV terminals can obtain more and more TV program content. However, such an excess of TV shows also burdens TV viewers, as it takes longer to search for their favorite TV show content. The recommendation system can help users to efficiently filter out the information they are interested in, and ...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to view more

Application Information

Patent Timeline
no application Login to view more
IPC IPC(8): G06F16/36G06F16/25G06F16/951G06F16/9535G06F40/30G06N3/04G06N3/08
CPCG06F16/367G06F16/258G06F16/951G06F16/9535G06F40/30G06N3/08G06N3/048
Inventor 詹会兰向超雷航杨茂林
Owner UNIV OF ELECTRONIC SCI & TECH OF CHINA
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap