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

Conference summary method and model based on improved TextRank algorithm

A meeting minutes and meeting technology, applied in the field of data processing, can solve the problems of poor readability of summaries, unsuitable for generating longer text summaries, unsuitable for meeting minutes generation scenarios, etc., to achieve the effect of ensuring legibility

Pending Publication Date: 2022-04-12
INSPUR TIANYUAN COMM INFORMATION SYST CO LTD
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0007] Third, based on algorithms that directly use generative text summaries to generate summaries, such as the TextSum model, although the summaries can be constructed through full-text ideas to solve the problem of content omissions to a certain extent, such algorithms are not suitable for generating longer text summaries. When generating longer text summaries, the summaries are less readable and not suitable for meeting minutes generation scenarios

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
  • Conference summary method and model based on improved TextRank algorithm
  • Conference summary method and model based on improved TextRank algorithm
  • Conference summary method and model based on improved TextRank algorithm

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0098] combined with figure 1 , the present embodiment proposes a method for meeting minutes based on the improved TextRank algorithm, and its implementation includes:

[0099] (1) On-site recording of the meeting content, and converting the recording content into meeting text by calling the real-time voice-to-text interface.

[0100] (2) Mark the key content of the meeting text, and store the marked key content separately.

[0101] (3) Sorting out the common words in the sentences of the key content in the meeting scene, and constructing a meeting scene keyword thesaurus. The meeting scene keyword thesaurus is not limited to first, second, need to be clear, most important, first, second.

[0102] (4) Based on the TextRank algorithm, the voting mechanism is used to sort the importance of the sentences contained in the meeting text. During the voting process, the sorting results are changed through the manually marked key content and the keyword vocabulary, and the TextRank al...

Embodiment 2

[0133] combined with figure 2 , the present embodiment proposes a meeting minutes model based on the improved TextRank algorithm, and its structure includes: a recording conversion module 1, a marking module 2, a building module 3, a TextRank algorithm module 5, a threshold setting module 6, a judgment module 7, and a direct output Module 8, screening output module 9.

[0134] Through the recording conversion module 1, the meeting content is recorded on-site, and the real-time voice-to-text interface is called to convert the recording content into the meeting text.

[0135] Through the marking module 2, the manual is assisted to mark the key content of the meeting text, and the marked key content is stored separately.

[0136] Through the construction module 3, sort out the common words of the sentences in the key content in the meeting scene, and build the meeting scene keyword thesaurus 4, the meeting scene keyword thesaurus 4 is not limited to first, second, need to be cl...

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 conference summary method based on an improved TextRank algorithm, and relates to the technical field of data processing, and the method comprises the steps: carrying out the on-site recording of conference content, and converting the recorded content into a conference text; marking key contents of the conference text; carding common words of statements where the key content is located, and constructing a keyword library; on the basis of a TextRank algorithm, statements contained in a conference text are subjected to importance sorting by utilizing a voting mechanism, in the voting process, a sorting result is changed through key contents marked manually and a keyword library, and the TextRank algorithm outputs abstract contents on the basis of the changed sorting result; and setting the number of statements contained in the conference summary, and directly or deleting the number of statements contained in the summary content output by the TextRank algorithm according to the sorting result to realize the automatic output of the final conference summary. The invention further discloses a conference summary model which is used for realizing the conference summary method, and the readability, the accuracy and the integrity of the output conference summary in different scenes can be ensured.

Description

technical field [0001] The invention relates to the technical field of data processing, in particular to a meeting minutes method and model based on an improved TextRank algorithm. Background technique [0002] The generation of traditional meeting minutes is mainly through manual summarization. This method requires a huge amount of manpower, has great limitations in timeliness, and has relatively subjective factors in the content of meeting minutes. [0003] In order to reduce the labor cost of writing meeting minutes, the current mainstream method is to automatically generate meeting minutes based on the idea of ​​text summarization. Representative text summarization algorithms include extractive-based TextRank, LexRank, TextTeaser, HeadTail and generative-based TextSum and other text summarization algorithms. Using the above text summarization algorithm, a text summarization is directly generated from the meeting text data and used as meeting minutes. [0004] At presen...

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 Applications(China)
IPC IPC(8): G06F16/34G06F40/289G10L15/26
Inventor 王凯袁明明李尚锴
Owner INSPUR TIANYUAN COMM INFORMATION SYST CO LTD
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