Application of Artificial Intelligence technology to the analysis and synthesis of reliable software systems

The development of a methodology for the production of highly reliable software is one of the greatest challenges facing the computer industry. Meeting this challenge will undoubtably involve the integration of many technologies. This paper describes the use of Artificial Intelligence technologies in the automated analysis of the formal algebraic specifications of abstract data types. These technologies include symbolic execution of specifications using techniques of automated deduction and machine learning through the use of examples. On-going research into the role of knowledge representation and problem solving in the process of developing software is also discussed.