Search-based program analysis

  • Authors:
  • Andreas Zeller

  • Affiliations:
  • Saarland University, Saarbrücken, Germany

  • Venue:
  • SSBSE'11 Proceedings of the Third international conference on Search based software engineering
  • Year:
  • 2011

Quantified Score

Hi-index 0.00

Visualization

Abstract

Traditionally, program analysis has been divided into two camps: Static techniques analyze code and safely determine what cannot happen; while dynamic techniques analyze executions to determine what actually has happened. While static analysis suffers from overapproximation, erring on whatever could happen, dynamic analysis suffers from underapproximation, ignoring what else could happen. In this talk, I suggest to systematically generate executions to enhance dynamic analysis, exploring and searching the space of software behavior. First results in fault localization and specification mining demonstrate the benefits of search-based analysis.