A General Algebraic Theory and Derived Software Framework for Automated Tradeoff Analysis Tools

Report
Authors:Tang, Chong, Department of Computer ScienceUniversity of Virginia Sullivan, Kevin, Department of Computer ScienceUniversity of Virginia Bagheri, Hamid, Department of Computer ScienceGeorge Mason University Dou, Ke, Department of Computer ScienceUniversity of Virginia
Abstract:

This paper presents a formal, general algebraic theory of design space tradeoff analysis tools, and a map-reduce-based framework, derived mechanically from the theory, for implementing such tools. The theory is organized as a hierarchy of Coq typeclasses in a style similar to that being used elsewhere to formalize abstract mathematics. From this theory, using Coq's extraction function, we produce a polymorphic framework (in Scala) that developers specialize and extend to produce domain-specific trade-off analysis tools. As a test and demonstration, we instantiated this framework using code re-engineered from an earlier, ad hoc, only partially automated system for tradeoff analysis of object-relation mappings. Our new tool reduced the time taken by one analysis from weeks to hours.

Rights:
All rights reserved (no additional license for public reuse)
Language:
English
Source Citation:

Tang, Chong, Kevin Sullivan, Hamid Bagheri, and Ke Dou. "A General Algebraic Theory and Derived Software Framework for Automated Tradeoff Analysis Tools." University of Virginia Dept. of Computer Science Tech Report (2014).

Publisher:
University of Virginia, Department of Computer Science
Published Date:
2014