A method and system for identify primary and secondary relations of Chinese text

A relationship recognition and text technology, applied in the field of text analysis, can solve the problem of ignoring the importance of text structure analysis, and achieve the effect of improving the recognition performance

Pending Publication Date: 2019-02-01
SUZHOU UNIV
View PDF1 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Most of the research on discourse primary and secondary relationship recognition focuses on the English corpus RST-DT. These studies generally regard discourse primary and secondary relationship recognition as an auxiliary link in discourse rhetorical structure analysis, ignoring its importance in discourse structure analysis

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 method and system for identify primary and secondary relations of Chinese text
  • A method and system for identify primary and secondary relations of Chinese text
  • A method and system for identify primary and secondary relations of Chinese text

Examples

Experimental program
Comparison scheme
Effect test

example 4

[0078]

[0079]

[0080] Among them, Sentence represents the text units contained in the primary and secondary relationship of the text, and the text units are separated by "|", and a primary and secondary relationship of a text contains two or more text units. Center indicates the primary and secondary relationship types of chapters, where "1" indicates the "core-satellite" relationship, "2" indicates the "satellite-core" relationship, and "3" indicates the "core-core" relationship. In this example, two primary and secondary relationship instances of chapters are marked.

[0081] The primary and secondary relationship of the chapters in Example 4 is expressed as:

example 5

[0083] "However, this kind of legal system closely following economic and social activities has been well received by domestic and foreign investors. They believe that there are rules and regulations when investing in Pudong New Area, and the interests can be guaranteed.","1"

[0084] "They believe that there are rules and regulations when investing in the Pudong New Area, | follow the rules, | interests can be guaranteed.","3"

[0085] In Example 5, the content in the first quotation mark represents the chapter unit, and the text units are separated by "|", and the content in the second quotation mark represents the primary and secondary relationship type of the chapter. Commas are used to separate chapter units and chapter primary-subordinate relationship types.

[0086] S102. According to the obtained set of primary and secondary relations of the articles, convert the non-binary primary and secondary relations of the articles into the binary primary and secondary relations ...

example 6

[0096] "They believe that there are rules and regulations when investing in Pudong New Area, | follow the rules,","3"

[0097] "They believe that when investing in Pudong New Area, there are rules and regulations, and the interests can be guaranteed.","3"

[0098] In Example 6, there are two binary primary and secondary relations of texts, that is, the expression form after transforming non-binary relations into binary relations, and each primary and secondary relation of texts contains only two textual units.

[0099] S20. Invoke a word segmentation tool for each text unit in the set of binary primary and secondary relations of texts to obtain the word features of the primary and secondary relations of each text, and generate a set of word features of the primary and secondary relations of binary texts. Call the part-of-speech extraction tool for each text unit in the feature set of primary and secondary relational words in a binary text to obtain the part-of-speech features ...

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

A method for identify that primary and secondary relations of Chinese text includes reading the text unit labeling information and the text primary-secondary relation type labeling information in thelabeling document set, obtaining the text primary-secondary relation set, converting the non-binary text primary-secondary relation in the text primary-secondary relation set to the left sub-tree, andobtaining the binary text primary-secondary relation set. The set of annotated documents is a set of documents that have been annotated with the text primary-secondary relationship type, and is the basis of the training model of the present invention. The present invention relates to a system and method for identifying text primary and secondary relations based on gated memory network, which converts predicted text units into word vectors, and automatically captures hidden features between text units and more important information relative to the whole by using gated memory neural network. Compared with the prior art method and system, the method and system of the present invention improve the performance of Chinese text primary-secondary relation recognition.

Description

technical field [0001] The invention relates to the technical field of discourse analysis, in particular to a method and system for identifying primary and secondary relations in Chinese discourses. Background technique [0002] Discourse refers to length and chapters, which are the research objects of natural language understanding. They are natural language texts formed through semantic association and structural organization. Discourses have cohesion, coherence, intentionality, and acceptability. Acceptability, informativity, situationality, and intertextuality are seven basic characteristics. Discourse analysis includes three subtasks: structure construction, primary and secondary identification, and relationship classification. Among them, the primary and secondary relationship of the text indicates the relationship between the primary and secondary content within the text. The main content refers to the dominant and decisive part of the text, while the secondary cont...

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 & AuthorityApplications(China)
IPC IPC(8): G06F17/21G06F17/27
CPCG06F40/117G06F40/221G06F40/289
Inventor王体爽李培峰朱巧明周国栋张玉华
OwnerSUZHOU UNIV