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

A Text Similarity Network Construction Method Based on Expert Voting

A text similarity network and construction method technology, applied in special data processing applications, instruments, electrical digital data processing, etc., can solve the problem that the global threshold method cannot accurately control the similarity, cannot reflect the characteristics of different links, and the global threshold method does not work. Can support the dynamic expansion of text similarity network and other problems

Inactive Publication Date: 2016-04-27
SHANGHAI UNIV
View PDF1 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0003] (1) The global threshold method cannot reflect the different link characteristics of different texts
[0004] (2) The global threshold method cannot be accurately controlled according to the similarity of the two texts involved in the link
[0005] (3) The global threshold method cannot support the dynamic expansion of the text similarity network. When adding new text, the global threshold needs to be recalculated

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 Text Similarity Network Construction Method Based on Expert Voting
  • A Text Similarity Network Construction Method Based on Expert Voting
  • A Text Similarity Network Construction Method Based on Expert Voting

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0022] Embodiment one: see figure 1 , the text similarity network construction method based on expert voting, is characterized in that the link between texts is precisely controlled by the local threshold value generated by the expert voting method, which reflects the different link characteristics between different texts, and supports the dynamic similarity network expand;

[0023] The local threshold is the similarity threshold for establishing a link between any two texts;

[0024] The expert voting method, whose local threshold The calculation formula is as follows:

[0025]

[0026]

[0027]

[0028] in, is the expert vote value for text i, is the similarity set between text i and other texts, for collection The sum of the similarities in for collection The maximum similarity in for collection The minimum similarity in for collection The number of similarities in It is the minimum value among the expert vote values ​​of text i and text j....

Embodiment 2

[0029] Embodiment 2: This text similarity network construction method based on expert voting uses 70 papers from TKDE from 2011 to 2012 to construct a text similarity network. Such as figure 1 As shown, a method for building a text similarity network based on expert voting in this embodiment, the steps are as follows:

[0030] S1. Input field anthology, for example, input 70 texts of TKDE;

[0031] S2. Text representation and similarity measurement, for example, using a graph structure-based text representation model and similarity measurement method;

[0032] S3. Use the expert voting method to establish links between texts; the expert voting method formula, its local threshold The calculation formula is as follows:

[0033]

[0034]

[0035]

[0036] in, is the expert vote value for text i, is the similarity set between text i and other texts, for collection The sum of the similarities in for collection The maximum similarity in for collection Th...

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 discloses a text similarity network construction method based on expert voting. The text similarity network construction method specifically comprises the following steps: (1) inputting a filed text set; (2) showing the text and measuring the similarity; (3) establishing interlinking among texts by using an expert voting method; and (4) outputting the text similarity network. According to the text similarity network construction method, the expert voting method is adopted to generate a local threshold for precisely controlling the interlinking among the texts, so that different interlinking characteristics of different texts are reflected, so that the dynamic extension of the similarity network is supported; and the method is simple and easy to operate, and has a good effect.

Description

technical field [0001] The invention relates to a method for constructing a text similarity network, in particular to using an expert voting method to determine the similarity threshold of whether a link is established between any two texts, and then establishing a text similarity network according to the local threshold, which is a method based on expert voting Text Similarity Network Construction Method. Background technique [0002] At present, the common method of constructing text similarity network is the global threshold method. The global threshold method is a method of setting the similarity threshold of all texts by artificial or machine learning methods, and then building a similar network of texts based on the global threshold, but this global threshold method has the following shortcomings: [0003] (1) The global threshold method cannot reflect the different link characteristics of different texts. [0004] (2) The global threshold method cannot perform preci...

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): G06F17/30G06F17/27
Inventor 陈雪吴超
Owner SHANGHAI 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

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com