Crowdsourced Bug Triaging: Leveraging Q&A Platforms for Bug Assignment

[Download Pdf]

Type: Publication (accepted)

Venue: 19th International Conference on Fundamental Approaches to Software Engineering (FASE)
April 2-8, 2016, Eindhoven, The Netherlands

Authors: Ali Sajedi Badashian, Abram Hindle, Eleni Stroulia
Department of Computing Science, University of Alberta, Canada

Abstract
Bug triaging, i.e., assigning a bug report to the “best” person to address it, involves identifying a list of developers that are qualified to understand and address the bug report, and then ranking them according to their expertise. Most research in this area examines the description of the bug report and the developers’ prior development and bug-fixing activities. In this paper, we propose a novel method that exploits a new source of evidence for the developers’ expertise, namely their contributions in Stack Overflow, the popular software Question and Answer (Q&A) platform. The key intuition of our method is that the questions a developer asks and answers in Stack Overflow, or more generally in software Q&A platforms, can potentially be an excellent indicator of his/her expertise. Motivated by this idea, our method uses the bug-report description as a guide for selecting relevant Stack Overflow contributions on the basis of which to identify developers with the necessary expertise to close the bug under examination. We evaluated this method in the context of the 20 largest GitHub projects, considering 7144 bug reports. Our results demonstrate that our method exhibits superior accuracy to other state-of-the-art methods.

Keywords:
Bug triaging, bug assignment, crowdsourcing, GitHub, Stack Overflow

Featured presentation; "Crowdsourced Bug Triaging"

Type: Presentation

Venue: Consortium for Software Engineering Research (CSER) Fall 2015, Markham, ON, Canada

I presented the recently published ICSME 2015 paper again in the CSER fall meeting (Sunday November 1st 2015) at Markham, Ontario.

Github's Big Data Adaptor: An Eclipse Plugin

[Download Pdf]

Type: Publication

Venue: Conference of the Centre for Advanced Studies on Collaborative Research (CASCON) 2015, November 2-4, Markham, ON, Canada

Authors: Ali Sajedi Badashian*, Vraj Shah**, Eleni Stroulia*
* Department of Computing Science, University of Alberta, Canada
** Indian Institute of Technology

Abstract
The data of GitHub, the most popular code-sharing platform, fits the characteristics of “big data” (Volume, Variety and Velocity). To facilitate studies on this huge GitHub data volume, the GHTorrent web-site publishes a MYSQL dump of (some) GitHub data quarterly. Unfortunately, developers using these published data dumps face challenges with respect to the time required to parse and ingest the data, the space required to store it, and the latency of their queries. To help address these challenges, we developed a data adaptor as an Eclipse plugin, which efficiently handles this dump. The plugin offers an interactive interface through which users can explore and select any field in any table. After extracting the data selected by the user, the parser exports it in easyto-use spreadsheets. We hope that using this plugin will facilitate further studies on the GitHub data as a whole.

Keywords:
GitHub, Mining software repositories, Eclipse Plugin, Software tools, Big data

Crowdsourced Bug Triaging

[Download Pdf]

Type: Publication

Venue: 31st International Conference on Software Maintenance and Evolution (ICSME2015)
Sep 29 - Oct 1, 2015, Bremen, Germany

Authors: Ali Sajedi Badashian, Abram Hindle, Eleni Stroulia
Department of Computing Science, University of Alberta, Canada

Abstract
Bug triaging and assignment is a time-consuming task in big projects. Most research in this area examines the developers’ prior development and bug-fixing activities in order to recognize their areas of expertise and assign to them relevant bug fixes. We propose a novel method that exploits a new source of evidence for the developers’ expertise, namely their contributions to Q&A platforms such as Stack Overflow. We evaluated this method in the context of the 20 largest GitHub projects, considering 7144 bug reports. Our results demonstrate that our method exhibits superior accuracy to other state-of-theart methods, and that future bug-assignment algorithms should consider exploring other sources of expertise, beyond the project’s version-control system and bug tracker.

Keywords:
Bug Triaging, Bug Assignment, Software Social Networks, Expertise Seeking, Github, Stack Overflow

AITF (Alberta Innovates - Technology Futures Graduate Student) Scholarship

Type: Scholarship

I was awarded the "AITF (Alberta Innovates - Technology Futures Graduate Student) Scholarship" at May 2015 for two years.

Featured presentation: Social Software Development

Type: Presentation

Venue: IBM CAS Research 2015 University Days; Smart Interactions and Services track,
IBM Canada Software Lab, Markham, ON, Canada

On behalf of my supervisor, Professor Eleni Stroulia, I presented an overview of our current research in the area od "Social Software Development".
The outline of the talk was as follows:

• Overview of current social platforms their commonalities and differences   
• Reflecting on the GitHub and StackOverflow communities 
• Some insights on how social perceptions of expertise can be used to support development activities