Dissertation proposal

Full title:

A Declarative Specification System for More Agile Measurement, Analysis, and Visualization Targeted at Software and Systems Engineering

Short title:

More Agile Measurement for Software and Systems

When:

Tuesday, July 14, 2:30pm

Where:

Room 2101 in the Collaborative Innovation Center (CIC 2101) on the Carnegie Mellon Campus

Committee:

  • Dr. William Scherlis, Carnegie Mellon
  • Dr. James Herbsleb, Carnegie Mellon
  • Dr. Anita Sarma, Carnegie Mellon
  • Dr. Barry Boehm, University of Southern California

Link to proposal document:

http://maccherone.com/publications/Proposal.pdf

Short abstract (longer abstract in proposal document):

There is ready agreement that better understanding (as in knowledge and situational awareness) generally leads to better decisions and outcomes. However, there are competing ways to gain better understanding. Unfortunately, for too large a number of important questions, developers rely upon intuition and tacit knowledge rather than more precise means including measurement, analysis, and visualization because the burden of collecting data is too high and the predictive power of things that can be readily measured is not strong enough. So, if we can drive down the cost and increase the predictive power of measurement we can increase the number of questions that are answered with more precise means. I propose a way to declaratively specify a measurement, analysis, and visualization system that capitalizes upon the following opportunities:
    • Passively acquiring data. The increasing richness and accessibility of data that is passively gathered makes an ad-hoc, just-in-time approach to measurement feasible.
    • Leveraging tacit knowledge. A big improvement to current measurement regimes can be achieved if it works in conjunction with tacit knowledge; allowing the user to explore the data, see the forest for the trees, rapidly confirm or refute hypotheses, devise new questions from the output of current analysis, and spiral in on the answers.
    • Rapidly iterating. The more measures (and visualizations) we can try, the more likely we are to find ones with greater predictive power (or better ability to express reality).
    • Considering coordination, dependencies, and relationships. Software development is done on interdependent components by individuals that have a variety of different relationships (social, organizational, geographic, etc.). If our measurement systems took these dependencies into account, they could have predictive power with respect to these important issues.
Automatically executing a measurement regime specified in this proposed syntax should be a significant improvement over traditional means. This dissertation will try to show that we can achieve this “more agile measurement.”
 

Comments are closed.