Kostadin Damevski, Ph.D.

Associate Professor | VCU College of Engineering

Richmond, VA, UNITED STATES

Interested in software engineering and in the use of natural language processing techniques to improve software maintenance and evolution.

Links (3)

Biography

Dr. Kostadin Damevski is an Associate Professor in the Department of Computer Science at Virginia Commonwealth University School of Engineering. His current research interests are centered around software maintenance and evolution, applied to domains such as mobile apps, high-performance computing, and industrial software systems. His research has been supported by U.S. government agencies, e.g., NSF, DARPA, DOE as well as private industry, e.g., Google, ABB. Damevski leads the Software Improvement (SWIM) Lab at VCU.

Industry Expertise (3)

Computer Software

Research

Education/Learning

Areas of Expertise (4)

Software Engineering

Software Maintenance

Recommendation Systems

Natural Language Processing

Education (1)

University of Utah: Ph.D., Computer Science 2007

Affiliations (1)

Associate Editor, IEEE Software

Selected Articles (2)

Fast changeset-based bug localization with BERT

44th International Conference on Software Engineering (ICSE 2020)

Agnieszka Ciborowska, Kostadin Damevski

2022-05-01

Helping developers to localize bugs (using bug reports) to the changeset that induced them. Changesets (or diffs) can be more useful for fixing bugs than static source code (e.g., methods or classes) as they encode the change that created the bug and include a (usually) meaningful message.

Changeset-Based Topic Modeling of Software Repositories

IEEE Transactions on Software Engineering

Christopher S. Corley, Kostadin Damevski, Nicholas A. Kraft

2019-06-01

The standard approach to applying text retrieval models to code repositories is to train models on documents representing program elements. However, code changes lead to model obsolescence and to the need to retrain the model from the latest snapshot. To address this, we previously introduced an approach that trains a model on documents representing changesets from a repository and demonstrated its feasibility for feature location. In this paper, we expand our work by investigating: a second task (developer identification), the effects of including different changeset parts in the model, the repository characteristics that affect the accuracy of our approach, and the effects of the time invariance assumption on evaluation results.