Patent · US Active

Modeling english sentences within a distributed neural network for comprehension and understanding of a news article

US10929609B1 · kind B1 · utility

2Cited by
0References
8Claims
0Family size

Assignee

Inventor

Key dates

Filing dateJun 26, 2018
Grant dateFeb 23, 2021
Priority date
Expiry dateAug 7, 2038

Classification

  • Technology area (CPC G)Physics
  • CPC primaryG06N3/105
  • WIPO fieldComputer technology
  • WIPO sectorElectrical engineering

Abstract

A distributed neural network comprehend news articles that emulate language processing of English sentences and derive basic understanding. The method of computation implements the MAYA Semantic Technique (Mitchell, 2015), a mathematical technique I previously developed to determine partial semantics of sentences within a natural language processing program. The results of the model shows how a distributed neural network that composed of the words, part-of-speech and its lexical synset can be constructed to comprehend news articles and determine basic understanding of the text. The text can be from either news, journals or magazine articles.

Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.