![]() |
College of Computing Sciences |
![]() |
Software Reuse Cost Estimation
In this paper (Annals of Software Engineering, 2001), we had proposed a software reuse cost model that has the following characteristics:
There are four stakeholders in the reuse process, namely: the corporate management, the domain engineering team, the application engineering team, and the component engineering teams.
Each stakeholder is responsible for a key decision in the process, and the collaboration of all stakeholders is necessary to ensure the proper conduct of reuse activities.
All stakeholder decisions are based on an economic rationale, and all can be modeled as investment decisions.
All stakeholder decisions can be quantified by means of return on investment functions, such as NPV, ROI, ARR, IRR, ARBV, PB, etc.
It is possible to optimize the conduct of the corporate reuse activity by setting controllable factors in such a way as to maximize the corporate ROI under the constraint that all ROI's are positive (or greater than a predefined threshold).
Our current research builds on this original model by proposing:
A product line of automated tools that provide corporate-wide support for decision making and cost estimation. Variabilities in this product line include variance in reuse organizations, variance in ROI functions, variance in database support (for archival data), variance in procurement channels, variance in cost models, and variance in access rights.
A function that automates the derivation of optimal reuse parameters by optimizing the corporate ROI while keeping all relevant ROI's positive (to create incentives for all relevant stakeholders).
A cost model that captures time to market (saved by reuse) in the cost equations.
This work is currently in progress. A thesis is expected to be completed on this topic in February 2008.