Automatically generating object models from natural language analysis

  • Authors:
  • Hector G. Perez-Gonzalez;Jugal K. Kalita

  • Affiliations:
  • Univ. of Colorado at Colo. Springs, Colorado Springs. CO;Univ. of Colorado at Colo. Springs, Colorado Springs. CO

  • Venue:
  • OOPSLA '02 Companion of the 17th annual ACM SIGPLAN conference on Object-oriented programming, systems, languages, and applications
  • Year:
  • 2002

Quantified Score

Hi-index 0.00

Visualization

Abstract

Our goal is to enable rapid production of static and dynamic object models from natural language description of problems. Rapid modeling is achieved through automation of analysis tasks. This automation captures the cognitive schemes analysts use to build their models of the world through the use of a precise methodology. The methodology is based on the use of proposed technique called role posets, and a semi-natural language (called 4W). Original problem statements are automatically translated to 4W language. The produced sentences then, are analyzed with role posets to produce static model views. Finally the 4W sentences are used to generate dynamic views of the problem. This set of methods maximizes analysis process agility, promotes reusability and constitutes a valuable tool in the learning process of object thinking.