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

Generative conference abstracting method based on graph convolutional neural network

A convolutional neural network and conference abstract technology, applied in the field of generative conference abstracts, which can solve the problem of ignoring the structural information of dialogue chapters

Active Publication Date: 2020-07-28
HARBIN INST OF TECH
View PDF4 Cites 7 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The purpose of the present invention is to solve the problem that the existing method only uses the sequence structure of sentences and words to model the conference text, ignoring the rich dialogue structure information of the conference, and proposes a generative meeting summary method that incorporates dialogue structure information

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
  • Generative conference abstracting method based on graph convolutional neural network
  • Generative conference abstracting method based on graph convolutional neural network
  • Generative conference abstracting method based on graph convolutional neural network

Examples

Experimental program
Comparison scheme
Effect test

specific Embodiment approach 1

[0028] Specific implementation mode one: combine figure 1 Describe this implementation mode. In this implementation mode, a generative conference summarization method based on graph convolutional neural network has a specific process as follows:

[0029] Step 1: Train the dialogue discourse structure parser, and mark the meeting data to obtain the dialogue discourse structure of the meeting;

[0030] Step 2: For each meeting in the meeting data set and the dialogue chapter structure of the meeting obtained in step 1, construct a meeting chapter structure diagram. The meeting chapter structure diagram includes all sentences in the meeting, all participants in the meeting, and Dialogue discourse structure between sentences;

[0031] Denote the conference chapter structure graph as G D =(V D ,E D , R D ), where v i ∈V D represent graph nodes, (v i ,r,v j )∈E D Represents the edge in the graph, r∈R D Represents the edge relationship in the graph;

[0032] Step 3: Use t...

specific Embodiment approach 2

[0038] Embodiment 2: This embodiment differs from Embodiment 1 in that: in the step 1, train the dialogue text structure parser, and mark the meeting data to obtain the dialogue text structure of the meeting; the specific process is:

[0039] Step 11. Use the existing STAC dataset [5] (Title: Discourse structure and dialogueacts in multiparty dialogue: The stac corpus, Authors: Nicholas Asher, Julie Hunter, Mathieu Morey, Farah Benamara, and Stergos Afantenos, Year: 2016) Training the existing dialogue structure parser Deep Sequential [6] (Title: Adeep sequential model for discourse parsing on multi-party dialogues, author: Zhouxing Shi and Minlie Huang, time: 2019, literature quoted from: Proceedings of the AAAI Conference on Artificial Intelligence), get the trained dialogue text structure parser Deep Sequential;

[0040] Step 1 and 2: Use the trained dialogue discourse structure parser Deep Sequential to mark the dialogue discourse structure of the AMI meeting data, and obt...

specific Embodiment approach 3

[0042] Specific embodiment three: what this embodiment is different from specific embodiment one or two is: utilize existing STAC data set to train existing dialogue discourse structure parser Deep Sequentia in the described step one by one, obtain the dialogue discourse structure analysis that has trained Deep Sequential; the specific process is:

[0043] The STAC dataset is a multiplayer chat dataset for English games. This corpus marks the semantic relationship between the basic semantic units (EDU) in the dialogue;

[0044] Said multiple persons are 3 or more persons.

[0045] According to statistics, STAC contains a total of 1091 dialogues, 10677 basic semantic units, and 11348 relational instances;

[0046] According to the task definition, at least 3 speakers participated in each dialogue. Each message sent by each speaker is usually 1-2 sentences each time, and each sentence is usually regarded as an elementary semantic unit (EDU) in multi-person dialogue discourse a...

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 generative conference abstracting method based on a graph convolutional neural network, and relates to the generative conference abstracting method based on the graph convolutional neural network. The method aims to solve the problem that according to an existing method, only sequence structures of sentences and words are used for modeling conference texts, and rich dialogue chapter structure information of a conference is ignored. The method comprises the steps of 1, obtaining a dialogue chapter structure of a conference; 2, constructing a conference chapter structure chart and a dialogue chapter structure between sentences in a conference; 3, constructing a conference chapter structure chart of the pseudo data and the corresponding pseudo data; 4, obtaining a pre-trained generative conference abstract model and initialization parameters of the graph neural network; obtaining a trained generative conference abstract model and model parameters of the graph neural network; and testing the conference to be tested by using the trained generative conference abstract model of the graph neural network to generate an abstract. The method is used for a generativeconference abstracting method in the field of natural language processing.

Description

technical field [0001] The present invention relates to a generative meeting summarization method based on graph convolutional neural network. Background technique [0002] Based on natural language processing - automatic text summarization (Automatic Summarization) [1] (Title: Constructing literature abstracts by computer: techniques and prospects, author: Chris D Paice, year: 1990, literature cited from Information Processing & Management) under the field of abstractive meeting summarization (Abstractive Meeting Summarization), that is, the text of a given multi-person meeting Record, generate a short text description containing key information of the meeting, as shown in Figure 2(a), which shows a meeting fragment and its corresponding standard summary. [0003] For conference summaries, existing methods can be divided into two categories: extractive and abstractive. The extraction method selects important sentences from the meeting minutes to form the final abstract. A...

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): G06F16/34G06N3/04G06N3/08
CPCG06F16/345G06N3/08G06N3/045
Inventor 冯骁骋秦兵冯夏冲刘挺
Owner HARBIN INST OF TECH
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