Technical Program
  Invited Speakers
  Technical Papers
  Practitioner Reports
Educators' Symposium
Doctoral Symposium
Student Research Comp.
Turing Lecture
Social Events
Week at a Glance
Final Program (1.5M .pdf)

Find in Program


view, help

"Legacy Transformation: Capturing Business Knowledge from Legacy Systems"
Object-Oriented Programming, Systems, Languages and Applications
Home    Program    Housing & Transportation    Registration    Submissions    Wiki    Maps

 : Posters on Display : Posters on Display : Posters on Display : Posters on Display at the Welcome Reception

Legacy Transformation: Capturing Business Knowledge from Legacy Systems

Jean B?zivin, Group (INRIA & LINA), University of Nantes
Sergio de Cesare, Brunel University
Grant Holland, Sun Microsystems
Mark Lycett, Brunel University
Chris Partridge, Brunel University

Legacy transformation is a significant problem that modern organizations currently need to cope with. Legacy systems are source of significant corporate knowledge (relating to concepts such as objects, rules, events, processes and services) which in many cases is not properly documented. The source code and data may represent the only source from which business models can be reverse engineered. Current transformation strategies are limited in number and mainly address the technological aspects. The primary concern of current approaches is migrating functionality and data from an obsolete technology platform to a new system developed with state-of-the art technology. Minor or no emphasis is given to the reverse engineering of business knowledge from the legacy systems and its representation in technology-agnostic business models. There is a growing need in the software development community for methods and techniques that address two complementary problems. Firstly, the reverse engineering of corporate/business knowledge from legacy systems in a traceable form. Secondly, being able to represent business requirements in models that can effectively trace to software models. From the perspective of Model Driven Architecture (MDA) solving this second problem would imply being able to represent computational independent models (CIM) and map these to platform independent models (PIM).