Check patentability & draft patents in minutes with Patsnap Eureka AI!

A Multi-Relational Question Answering System Using Explainable Reasoning Networks

A relational and network technology, applied in the fields of artificial intelligence and the Internet, can solve problems such as the lack of explicit explanations of neural interrogators and unsuitability for other operations, and achieve good performance and high interpretability

Active Publication Date: 2022-07-05
TSINGHUA UNIV
View PDF4 Cites 0 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, the Neural Programmer requires predefined symbolic operations and thus may not be suitable for other operations, while the Neural Interrogator lacks explicit interpretation

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 Multi-Relational Question Answering System Using Explainable Reasoning Networks
  • A Multi-Relational Question Answering System Using Explainable Reasoning Networks
  • A Multi-Relational Question Answering System Using Explainable Reasoning Networks

Examples

Experimental program
Comparison scheme
Effect test

Embodiment

[0041]First of all, it should be noted that, although Chinese or English are used as examples of natural languages ​​below, this is only for ease of understanding and does not imply any limitation. The applicable occasions of the present invention are not limited to specific language types or specific language families.

[0042] Task Definition: The goal is to provide an interpretable reasoning network to answer multi-relational questions.

[0043] Given a question q, its named entity or topic e s It can be obtained with NER (Named Entity Recognition) tools, the task is to find an entity a in the knowledge base as an answer.

[0044] The inventors considered two typical multi-relational problems: the path problem and the connection problem, and the former is the main focus of the present invention.

[0045] The path problem contains only named entities (body e s ) to find its answer (object a) along an answer path consisting of several relations and corresponding intermedia...

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 present application discloses a multi-relational question answering method using an interpretable reasoning network, which is carried out in a multi-hop manner. The method includes: step 1, converting an input question into a distributed question representation; The reasoning state of the jump and the problem of the previous jump indicate that the relationship of the current jump is predicted; step 3, if the predicted relationship of the current jump is not a termination relationship, then according to the predicted relationship of the current jump, the relationship of the previous jump is calculated. The reasoning state and the problem representation of the previous jump are updated to the reasoning state of the current jump and the problem representation of the current jump, and return to step 2 as the next jump, otherwise, go to step 4; and the step 4, according to the current The reasoning state of the jump, determine and output the answer of the current jump. The method provides traceable and observable intermediate predictions for inference analysis and fault diagnosis.

Description

technical field [0001] The present invention relates to the field of artificial intelligence and Internet technology, and in particular, to an (open domain) multi-relational question answering system using an interpretable reasoning network. Background technique [0002] The question answering system based on knowledge base is an advanced form of information retrieval system, which can use accurate and concise natural language to answer users' question and answer query service requests expressed in natural language. Specifically, given a natural language question, through semantic understanding and parsing of the question, the knowledge base is used to query, and the answer to the question is given by retrieving the relevant domain knowledge database. [0003] Among such question answering systems, open-domain question answering (QA) has been a hot topic in the field of artificial intelligence, and large knowledge bases such as Freebase have also facilitated this task in rec...

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 Patents(China)
IPC IPC(8): G06F16/332G06F16/36G06F16/33G06F40/295G06F40/284G06N5/04
CPCG06F16/3329
Inventor 黄民烈周曼桐朱小燕
Owner TSINGHUA UNIV
Features
  • R&D
  • Intellectual Property
  • Life Sciences
  • Materials
  • Tech Scout
Why Patsnap Eureka
  • Unparalleled Data Quality
  • Higher Quality Content
  • 60% Fewer Hallucinations
Social media
Patsnap Eureka Blog
Learn More

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

© 2025 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com