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

Answering complex questions by neural machine reading understanding dependent on utterance analysis

A technique for complex questions, discourses, applied in the field of generating or verifying answers to questions, capable of solving problems that cannot be answered reliably and accurately

Pending Publication Date: 2022-05-03
ORACLE INT CORP
View PDF0 Cites 1 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

But existing techniques cannot reliably and accurately answer the question

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
  • Answering complex questions by neural machine reading understanding dependent on utterance analysis
  • Answering complex questions by neural machine reading understanding dependent on utterance analysis
  • Answering complex questions by neural machine reading understanding dependent on utterance analysis

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0023] Aspects of the present disclosure relate to machine reading comprehension (MRC), which employs semantic analysis, syntactic analysis, general graph alignment, and discourse analysis to generate answers to questions. In particular, the disclosed techniques relate to generating and / or verifying representations (e.g., abstract meaning representations (AMRs), entity-based graph representations of text) and / or utterance representations (e.g., The answer to the question represented by the discourse tree). These techniques can be used to corroborate and / or correct answers generated by deep learning systems.

[0024] Combining knowledge about the target passage from sources such as syntactic parse trees (also referred to as "syntax trees" for brevity), semantic abstract meaning representation (AMR) parse results, and discourse trees can provide useful information for answering attribute-value questions. Tool of.

[0025] For machine reading comprehension (MRC), it is helpful ...

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 disclosure relates to answering complex questions dependent on utterance analysis by neural machine reading understanding. Specifically, an autonomous agent receives a user query that includes a complex question. The agent may obtain answer candidate text from a corpus of unstructured text, the answer candidate text corresponding to the user query and including text from which an answer is subsequently identified. The agent may generate first language data corresponding to the user query and second language data corresponding to the answer candidate text. Each instance of language data may include a respective combination of syntactic data, semantic data, and utterance data generated from user query and / or answer candidate text. Two instances of language data may be provided to a machine learning model, which has been previously trained to output answers identified from instances of unstructured text (e.g., answer candidate text). The model may output answers identified from the answer candidate text, which in turn may be provided in response to the user query.

Description

[0001] Cross References to Related Applications [0002] This application claims the benefit of U.S. Provisional Application No. 63 / 107,189, filed October 29, 2020, entitled "Relying on Discourse Analysis to Answer Complex Questions by Neural Machine Reading Comprehension," the contents of which are incorporated for all purposes In this article. technical field [0003] This disclosure relates generally to linguistics. More specifically, the present disclosure relates to using discourse analysis and neural machine reading comprehension to generate or validate answers to questions. Background technique [0004] Computer-implemented applications of linguistics are increasing due to enormous increases in processor speed and memory capacity. For example, computer-based linguistic discourse analysis facilitates numerous applications, such as automated agents that can answer questions received from user devices. But existing techniques cannot reliably and accurately answer the ...

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): G06F40/30G06F40/295G06F40/211G06N3/04G06N3/08G06N20/00
CPCG06F40/30G06F40/295G06F40/211G06N3/04G06N3/08G06N20/00G06F40/35G06F40/216G06F40/284G06F16/3329G06F16/3344G06N3/045G06N5/04
Inventor B·加利茨基
Owner ORACLE INT CORP
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