Eureka AIR delivers breakthrough ideas for toughest innovation challenges, trusted by R&D personnel around the world.

Interpretable reasoning question-answering method and device

A technology of questions and answers, applied in the field of interpretable reasoning question answering methods and devices, which can solve problems that affect the answer accuracy of the question answering system, insufficient completeness of the knowledge map, and reduce user experience, etc.

Pending Publication Date: 2021-04-02
厦门渊亭信息科技有限公司
View PDF9 Cites 6 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

In addition, in the process of information extraction, due to the complexity of data sources, the possibility of knowledge errors is high, and the information itself is difficult to obtain using the existing open data on the Internet. This situation makes the knowledge graph not complete enough and affects question answering. The accuracy of the system's answers reduces the user experience

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
  • Interpretable reasoning question-answering method and device
  • Interpretable reasoning question-answering method and device

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0046] refer to figure 1 , this embodiment provides an interpretable reasoning question answering method, including:

[0047] According to the question to be answered, construct a knowledge map of the field related to the question;

[0048] Converting the knowledge map into a graph structure, and predicting the semantic relationship between entities in the graph structure;

[0049] Obtaining logical rules between entities based on the semantic relationship;

[0050] Complementing the knowledge graph based on the logic rules;

[0051] According to the question to be answered, based on the completed knowledge graph, the best answer is deduced in the knowledge graph, and the answer to the question is obtained.

[0052] Aiming at the incompleteness of the existing knowledge graph, this disclosure studies the potential semantic and structural features between knowledge entities in the knowledge graph, and then mines the logical rules between entities, and uses the logical rules ...

Embodiment 2

[0095] refer to figure 2 , this embodiment provides an interpretable reasoning question answering device, including:

[0096] The knowledge map building module is used to construct the knowledge map of the field related to the question according to the question to be answered; the knowledge map building module includes knowledge extraction, knowledge representation and knowledge fusion. Knowledge extraction is further divided into: unstructured data cleaning, named entity recognition, and relationship extraction. When building Chinese knowledge graphs, this system uses a crawler framework for data processing, and an end-to-end model based on sequences and tree structures for knowledge extraction. After knowledge fusion, a visual graph database is used to store knowledge triples.

[0097] A query reasoning module, configured to mine logical rules between entities in the knowledge graph, and use the logical rules to complete the knowledge graph;

[0098]The natural language q...

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 provides an interpretable reasoning question-answering method and device. The method comprises the steps of constructing a knowledge graph of related fields of a to-be-answered question;converting the knowledge graph into a graph structure, and predicting a semantic relationship between entities in the graph structure; based on the semantic relationship, obtaining a logic rule between entities; complementing the knowledge graph based on the logic rule; and according to the to-be-answered question, obtaining an answer to the question based on the completed knowledge graph. Relational reasoning is carried out by applying rules, and the goal of knowledge graph completion is achieved, so that the interpretability of the reasoning process is improved, a question-answering systemanswers more accurately, and the user experience is improved.

Description

technical field [0001] The invention relates to a question answering method, in particular to an interpretable reasoning question answering method and device. Background technique [0002] Intelligent answering systems have been a hot topic in the field of Al, and the recent development of numerous knowledge bases has facilitated the realization of this task. However, due to the diversity and complexity of languages ​​and knowledge, reasoning based on knowledge base question answering is still a challenging task, and knowledge graph based question answering systems are attracting more and more attention. The completeness and correctness of the knowledge graph are also becoming more and more important. Although the current knowledge graph contains hundreds of millions of facts, the knowledge in the knowledge graph is far from complete. For example, in the large-scale knowledge base Freebase, the basic information of more than 90% of the characters is incomplete, and some mi...

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): G06F16/36G06F16/332G06F16/35G06F40/279G06F40/30G06N3/04G06N5/04
CPCG06F16/367G06F16/3329G06F16/35G06F40/279G06F40/30G06N5/04G06N3/044
Inventor 洪万福钱智毅王旺
Owner 厦门渊亭信息科技有限公司
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
Eureka Blog
Learn More
PatSnap group products