Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

A Social Tag Recommendation Method Based on Random Walk and Conditional Random Field

A technology of conditional random field and random walk, which is applied in the direction of using information identifiers to retrieve web data, instruments, calculations, etc. It can solve the problems of reduced quality of tag recommendation, different levels of resource understanding and preference, and tag information noise, etc., to reduce Low-quality recommendations, quality-improved results

Active Publication Date: 2020-06-09
JIANGSU UNIV
View PDF1 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Most of the existing tag recommendation methods convert the three-dimensional relationship of social tags into two-dimensional relationships, and directly apply the traditional recommendation model, resulting in the underutilization of the three-dimensional relationship of user resource tags in the tag recommendation process, resulting in a decrease in the quality of tag recommendation
A resource is often tagged with multiple tags by multiple users. However, due to the difference in user understanding and preference of the resource, the tag information contains noise. Impact on labels

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
  • A Social Tag Recommendation Method Based on Random Walk and Conditional Random Field
  • A Social Tag Recommendation Method Based on Random Walk and Conditional Random Field
  • A Social Tag Recommendation Method Based on Random Walk and Conditional Random Field

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0042] In order to make the technical problems, technical solutions and advantages to be solved by the present invention clearer, the following will describe in detail with reference to the drawings and specific embodiments. The present invention selects the Cleaned Dump data set, which is the public data of the Bibsonomy website.

[0043] Extract the user label and resource label data in the data set to form the user label data set D 1, resource label dataset D 2 , the extraction process is as figure 2 shown. Select the feature vectors that can construct candidate tags for resources in the resource tag data set, including url URL information, description URL description information, titie book name, and description book description information. The dataset is described as D 2 ={(x i ,y i )|1≤i≤m}, where x i represents resources, y i Represents the tag added by the resource, where x i =[ulr, des_bm, title, des_bt] represents the 4-dimensional feature vector of resour...

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 present invention proposes a social tag recommendation method based on the random walk and the conditional random field. The method is characterized in that: by performing data extraction on tag recommendation data, user tag and resource tag data are extracted; a multi-tag classification algorithm based on the random walk is used to predict the tag probability that the resource may be annotated, and the characteristics of the resource itself are effectively used in the process, so that the influence of tag noise caused by user annotation is reduced; and the point feature function is designed by predicting the tag probability that the resource may be annotated, the edge feature function is designed by predicting the tag co-occurrence rate and the user similarity, the conditional randomfield is constructed to obtain the tag that the user may add for the resource, the tag greater than the threshold is taken as the recommended tag, and the three-dimensional relationship of the user, the resource, and the tag are effectively used in the process, so that the loss of effective information caused by converting the three-dimensional relationship of the social tag recommendation to thetwo-dimensional relationship is reduced. According to the method provided by the present invention, the quality of tag recommendation is improved, and the user can better find and locate social resources in a facilitated manner.

Description

technical field [0001] The invention relates to the application field of data mining and recommendation system, in particular to a social tag recommendation method based on random walk and conditional random field. Background technique [0002] The rapid development of information technology has brought about the problem of information overload, which has brought a heavy burden to users' information selection. In order to solve the problem of information overload, recommender system methods have been introduced. With the development of Web2.0 technology, more and more network users can freely choose tags to mark network resources, forming social tags. Tags can not only describe resources, but also express user preferences for resources. [0003] In a tag recommendation system, the task of social tag recommendation is to recommend a series of tags that the user may add to the resource given a user and a resource. Most of the existing tag recommendation methods convert the ...

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
Patent Type & Authority Patents(China)
IPC IPC(8): G06F16/9535G06F16/955
CPCG06F16/9535G06F16/955
Inventor 薛安荣夏欢欢曹静
Owner JIANGSU UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products