Jeff Tian's Research in Software Measurement and Analysis
I am interested in the measurement, analysis, and improvement of
software products and processes.
Some of my specific research activities in this area are summarized below.
Analysis of Software Defects and Other Data from Industry
Various measurement data about software processes and products
from industry can be analyzed to understand practical problems
and derive solutions to help optimize project management,
improve quality and productivity,
reduce cost and cycle-time.
Various statistical analysis techniques
(see also risk identification/management techniques)
can be used to understand the relations among different data attributes
and guide improvement actions.
Recent results for several projects from IBM and Nortel
are summarized in the following papers
(see also abstracts for some of
these and other recent papers):
-
A.G. Koru and J. Tian,
"An Empirical Comparison and Characterization of High Defect and High
Complexity Modules",
Journal of Systems and Software,
(accepted 4/2002, to appear.)
-
J. Tian, A. Nguyen, C. Allen, and R. Appan,
"Experience with Identifying and Characterizing Problem Prone Modules in
Telecommunication Software Systems",
Journal of Systems and Software,
Vol. 57, No.3, pp.207-215, July, 2001.
-
J. Tian and J. Troster.
"A Comparison of Measurement and Defect Characteristics of
New and Legacy Software Systems",
Journal of Systems and Software
Vol. 44, No. 2, pp.135-146, Dec., 1998.
Recently,
we have also been applying similar analysis techniques to analyze
the
defects and related data for web applications,
and obtained some interesting results.
Metrics Evaluation/Selection Criteria
I am interested in developing program complexity metrics based on formal
models (for example, decomposition of prime programs and information theory),
and models/techniques for metrics evaluation and selection,
to provide a solid basis for a successful measurement program.
This approach has been used in
NASA/SEL
projects and found to be effective.
This work is summarized in the following papers
(see also online abstracts
for these and other related papers):
-
J. Tian and M.V. Zelkowitz.
"Complexity Measure Evaluation and Selection",
IEEE Trans. on Software Engineering.
Vol.21, No.8, pp.641-650, Aug., 1995.
-
M.V. Zelkowitz and J. Tian.
"Measuring Prime Program Complexity".
Information Sciences.
Vol.77, No.3-4, pp.325-350, March, 1994.
Tool Support to Implement the above Techniques and Strategies
Some of my recent results in this research area are summarized in the
following paper:
-
J. Tian, J. Troster and J. Palma.
"Tool Support for Software Measurement, Analysis, and Improvement",
Journal of Systems and Software.
Vol. 39, No. 2, pp.165-178, 1997.
Prepared by Jeff Tian
(tian@engr.smu.edu).
Last update: May 2, 2003.
Back to Jeff Tian's home page