Distinguishing Fact from Opinion and Events from Meta-Events,

Abstract

A major problem in automatically analyzing the text of military messages in order to synthesize data base elements is separating fact from opinion, i.e., identifying factual statements versus evaluate commentary such as degree of belief or confidence in the parameters of a particular event, comparative evaluations of entitles and events, predictions of future events, etc. This paper describes a model which accounts for the range of factual to evaluative information in the message traffic, and discusses a means for representing such information in the context of an experimental system for automated data base generation.

Document Details

Document Type
Technical Report
Publication Date
Feb 01, 1983
Accession Number
ADP001167

Entities

People

  • Christine A . Montgomery

Tags

DTIC Thesaurus Topics

  • California
  • Computer Languages
  • Databases
  • Demographic Cohorts
  • Formal Languages
  • Language
  • Natural Language Processing
  • Natural Languages
  • Test And Evaluation

Readers

  • Artificial Intelligence
  • Systems Analysis and Design