Process knowledge extraction method and system for interactive dialogue

A knowledge extraction and knowledge technology, applied in the fields of unstructured text data retrieval, instruments, electronic digital data processing, etc., can solve the problems of fixed extraction process, inability to extract complex dialogues, and homogenization of extracted content, so as to avoid logic wrong effect

Pending Publication Date: 2022-04-12
AISPEECH CO LTD
View PDF0 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0009] In order to at least solve the problems of homogenization of extracted content, fixed extraction process, mismatch between knowledge logic and intent, and inability to extract complex dialogues in the existing technology

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
  • Process knowledge extraction method and system for interactive dialogue
  • Process knowledge extraction method and system for interactive dialogue
  • Process knowledge extraction method and system for interactive dialogue

Examples

Experimental program
Comparison scheme
Effect test

Embodiment approach

[0036] As an implementation manner, the abstracting the knowledge entity into a corresponding knowledge concept includes:

[0037] Based on entity clustering or entity type judgment, the knowledge entity is abstracted into a corresponding knowledge concept, wherein the degree of abstraction of the knowledge concept is inversely proportional to the process complexity of the interactive dialogue.

[0038] In this embodiment, the abstraction method may be entity clustering or entity type. For tasks in different scenarios, the degree of abstraction is also different. The more knowledge-intensive and complex the task, the lower the degree of abstraction. For simple tasks, maybe E 1 with E 2 The categories of these two similar entities can be regarded as a common knowledge concept C 1(such as "fever" and "fever"). For complex tasks, maybe E 1 There is no entity with which it can share the process, then E 1 is equivalent to concept C 1 .

[0039] For step S12, for each dialo...

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 embodiment of the invention provides a process knowledge extraction method for interactive conversations. The method comprises the following steps: determining a knowledge entity of each dialogue statement in an interactive dialogue by utilizing a knowledge graph, and abstracting the knowledge entities into corresponding knowledge concepts; performing intention extraction on each dialogue statement, and combining with the knowledge concept of each dialogue statement to obtain a plurality of sentence-level knowledge concept-intention combinations in the interactive dialogue; according to the dialogue sequence of the interactive dialogue, sequence training is conducted on the multiple knowledge concept-intention combinations based on the first knowledge concept-intention combination, and a knowledge concept-intention combination set conforming to the interactive dialogue process sequence is obtained. The embodiment of the invention further provides a process knowledge extraction system for the interactive dialogue. According to the embodiment of the invention, the feature combination of knowledge and intention is extracted and adopted, and the extracted knowledge is more in line with the content of the dialogue process. And extracting an abstraction process which is highly related to dialogue knowledge and has differences.

Description

technical field [0001] The invention relates to the field of intelligent speech, in particular to a process knowledge extraction method and system for interactive dialogue. Background technique [0002] In intelligent voice interaction, Q&A conversations are usually guided and interacted step by step by a preset process, so as to realize the service to users. Whether the dialogue process is set up humanized and rational, and whether the voice input by the user can be accurately mapped to the dialogue process will affect the user experience. In order to establish a humanized and rationalized dialogue process, it is necessary to extract the real dialogue process from the manual dialogue as a reference. [0003] Usually, extracting the real dialogue process from the artificial dialogue will use the feature extraction algorithm to characterize the question information in the dialogue to obtain the feature vector corresponding to the sentence. Based on the clustering method, si...

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): G06F16/36G06F40/295
Inventor 施淼元缪庆亮李茂龙杨一帆俞凯
Owner AISPEECH CO LTD
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

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

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap