IDEI 2286/2008

REASEARCHES ON ADAPTIVE OPTIMIZATION OF SOFTWARE SYSTEMS USING MACHINE LEARNING TECHNIQUES AND MULTIAGENT SYSTEMS

RO|EN

THE DEVELOPMENT OF TOOLS FOR BUILDING PERFORMANT, OPEN, SCALABLE AND CONTINUOUSLY ADAPTABLE SOFTWARE SYSTEMS AND THE DEVELOPMENT OF ARTIFICIAL INTELLIGENCE METHODS AND SYSTEMS ARE MAIN OBJECTIVES OF THE NATIONAL AND INTERNATIONAL RESEARCH. THE MAIN CAUSE OF MANY SOFTWARE PROJECTS FAILURE IS THE GREAT INFLUENCE OF THE HUMAN FACTOR AND THE LACK OF THEORETICAL MODELS. IN ORDER TO OVERCOME THESE DISADVANTAGES, MANY SOFTWARE ENGINEERING PROCESSES MUST BE AUTOMATED AND SOME ADEQUATE MATHEMATICAL MODELS MUST BE ELABORATED. THE PROBLEMS RELATED TO THE MAINTENANCE AND THE EVOLUTION OF SOFTWARE SYSTEMS ARE ESSENTIAL, A MAJOR REQUIREMENT BEING TO DEVELOP METHODOLOGIES FOR THEIR ADAPTIVE OPTIMIZATION (BOTH STRUCTURAL AND BEHAVIORAL). THE MAIN OBJECTIVE OF THIS REASEARCH PROJECT IS THE DEVELOPMENT OF MACHINE LEARNING METHODS AND MODELS FOR STRUCTURAL AND BEHAVIORAL SELF-ADAPTATION OF SOFTWARE SYSTEMS DURING THEIR MAINTENANCE AND EVOLUTION. WE WILL FOCUS OUR RESEARCHES ON DEVELOPING FORMAL MODELS FOR THE ESSENTIAL PROBLEMS THAT OCCUR IN TWO SUBFIELS OF SOFTWARE ENGINEERING, PROGRAM COMPREHENSION AND SOFTWARE REEINGINEERING, SUBFIELDS WHICH DEAL WITH PROBLEMS RELATED TO THE MAINTENANCE AND EVOLUTION OF SOFTWARE SYSTEMS. FOR DEVELOPING THE ADAPTIVE OPTIMIZATION METHODOLOGIES WE WILL USE MACHINE LEARNING TECHNIQUES. IN ORDER TO VALIDATE THE OBTAINED THEORETICAL RESULTS, WE AIM AT DEVELOPING AN INTELLIGENT MULTIAGENT DECISION SUPPORT SYSTEM FOR ASSISTING SOFTWARE DEVELOPERS IN THE MAINTENANCE AND EVOLUTION OF SOFTWARE SYSTEMS. WE INTEND TO FORM AN INTERDISCIPLINARY RESEARCH GROUP THAT WILL CONTINUE THE COLLABORATION WITH THE SEVERE RESEARCH GROUP FROM DETROIT. THIS WAY, WE WILL PROVIDE A SCIENTIFIC FRAMEWORK FOR THE YOUNG RESEARCHERS INVOLVED IN THIS PROJECT. OUR MAIN GOAL IS TO OBTAIN EXCELLENT, INTERNATIONALLY COMPETITIVE SCIENTIFIC RESULTS.

Copyright 2008, UBB. All rights reserved.