AbstFinder, A Prototype Natural Language Text Abstraction Finder for Use in Requirements Elicitation

  • Authors:
  • Leah Goldin;Daniel M. Berry

  • Affiliations:
  • Faculty of Computer Science, Technion, Haifa 32000, lsrael/ dberry@cs.technion.ac.il;Faculty of Computer Science, Technion, Haifa 32000, lsrael/ dberry@cs.technion.ac.il

  • Venue:
  • Automated Software Engineering
  • Year:
  • 1997

Quantified Score

Hi-index 0.00

Visualization

Abstract

Abstraction identification is named as a key problem in requirementsanalysis. Typically, the abstrac-tions must be found among the large mass ofnatural language text collected from the clients and users. This papermotivates and describes a new approach, based on traditional signalprocessing methods. for finding abstractions in natural language text andoffers a new tool, AbstFinder as an implementation of this approach. Theadvantages and disadvantages of the approach and the design of the tool arediscussed in detail. Various scenarios for use of the tool are offered. Someof these scenarios were used in case study of the effectiveness of rhe toolon an industrial-strength example of finding abstractions in a request forproposals.