Date of Award


Degree Type


Degree Name

Master of Science (MS)


Computer Science


Software engineering as practiced today (especially in the industry) is no longer about the stereotypical monolithic life cycle processes (e.g. waterfall, spiral, etc.) found in most software engineering textbooks. These heavyweight methods historically have impeded progress for small/medium sized development teams owing to their inherent complexity and rather limited data collection strategies that predominated the 1980s until relatively recently in the mid-2000s. The discipline and practice of software engineering includes software quality, which has an established theoretical foundation for doing software metrics. Software metrics are a critical tool which provide continuous insight to products and processes and help build reliable software in mission critical environments. Using software metrics we can perform calculations that help assess the effectiveness of the underlying software or process. The type of metrics relevant to our work are in-process metrics. In-process metrics focus on a higher-level view of software quality, measuring information that can provide insight into the underlying software development process.

In this thesis, we aim to develop and evaluate a metrics dashboard to support Computational Science and Engineering (CSE) software development projects. This task requires us to perform the following activities:

Assess how metrics are used and which general classes/types of metrics will be useful in CSE projects.

Develop a metrics dashboard that will work for teams using sites like Github, Bitbucket etc.

Assess the effectiveness of the dashboard in terms of project success and developer attitude towards metrics and process.

The challenge is to cherry-pick the most-effective practices from a large suite of tools and incorporate them into existing cloud-based workflows. As part of this thesis, we have developed a metrics dashboard based on the currently identified metrics types. The tools we are developing can be incorporated in any existing GitHub-based workflow without requiring the developers to install anything.

Creative Commons License

Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 License.