Finding Optimal Plans for Incremental Method Engineering

Abstract

Incremental method engineering proposes to evolve the information systems development methods of a software company through a step-wise improvement process. In practice, this approach proved to be effective for reducing the risks of failure while introducing method changes. However, little attention has been paid to the important problem of identifying an adequate plan for implementing the changes in the company's context. To overcome this deficiency, we propose an approach that assists analysts by suggesting-via automated reasoning-optimal and quasi-optimal plans for implementing method changes. After formalizing the Process-Deliverable Diagrams language for describing the method changes to implement, we present a planning framework for generating plans that comply with different types of constraints. We also describe an implementation of the modeling and planning components of our approach.

Method Association Approach: Situational construction and evaluation of an implementation method for software products

“…The method fragments of existing methods in our approach are modeled as Process Deliverable Diagrams (PDD) [23]. A tool as the one developed in the incremental method engineering approach [24] could be used to automatically convert the PDDs into logic programs in disjunctive Datalog that generate optimal plans.…”

“…As this step is independent from the others and as we needed method fragment to build the implementation software method for the case study, we decided to carry out this activity at this time of the study. The selected methods are the Unified Process [21], MOOSAD [22], SSA OnePoint Implementation Methodology (OPIM) -recently renamed Infor® Deployment Method , SDM [23] and GSDLC [24] as they include the notion of software implementation. We also included the already existing VitalHealth method in the HCM organization.…”