Jeff Tian's Applied Research in OO Systems

Specific Aspects of OO Systems

• Selection and application of appropriate metrics for OO systems.

  Several of the products from our industrial partners we analyzed have an important OO component but were all measured using traditional complexity and other metrics. We are in the process of selecting and using OO-specific metrics, which would, hopefully, lead use to better understandings of these systems. Similarly, some of the open source products we are measuring are OO systems. We should be able to get some interesting results.

• Selection and application of appropriate testing techniques for OO systems. This work is described in the following paper:

  • J. Tian, "Testing and Ensuring Reliability for Large OO Software Systems -- A Markov Usage Model and Data Partitioning Based Approach", OOPSLA'99-W12 Workshop on Modeling Reliability and Maintainability of Large-Scale Object-Oriented Software Systems, Denver, Colorado, Nov., 1999.

• Selection and application of other quality assurance (QA) techniques for OO systems. In particular, some approach based on formal methods, design pattern, and behavioral simulation seems promising. This work is described in the following papers:

  • M. Deshpande, F. Coyle and J. Tian, "Improving the Reliability of Object-Oriented Software through Object-level Behavioral Simulation", 5th International Conference on Reliable Software Technologies, pp.266-279, Potsdam (Berlin), Germany, June 2000.
    The proceedings was published as: H.B. Keller and E. Plodereder (Eds.), LNCS 1845: Reliable Software Technologies Ada-Europe 2000, Springer, 2000.

CBSE: Reliability Composition and Optimization

• The starting point is a good understanding and quantitative assessment of CBSE and how it affects system reliability, both positively (via using standard and high-quality components) and negatively (problem/failure propagation and magnification via reused components).
• Reliability composition: how to determine system reliability based on component reliability and system architecture.
• Reliability optimization: how to achieve maximal levels of reliability under given constraints, based on our understanding of the previous two problems, and system optimization techniques.