Empowering Open Source Software Communities

Supporting OSS Contributors in Issue Discussions

Open-source software development relies largely on discussions conducted in issue tracking systems such as GitHub Issues. Contributors exchange ideas, propose solutions, discuss approaches, and justify their decisions through comments. But as discussions get longer and involve more users, it gets harder for new contributors to understand how arguments relate, identify conflicting viewpoints, and follow the decision-making process. This can make collaboration difficult and longer to fix problems.

In this project, we focuse on facilitating understanding and navigation in open-source issue discussions. We aim to help new contributors better interpret complex discussions through analyzing argumentation structures and utilizing summarization techniques. Building on our prior projects understanding and supporting OSS communities, we are currently working on the design of a system that analyzes issue discussions, extracts the argumentative structure from issue posts and comments, and identifies the stance relationships between comments and the issue post, as well as between the comments themselves. In addition, the system supports contributors in writing more effective comments by providing feedback and assisting them in formulating clearer and better-structured arguments.

Fostering Developer Empathy in Open Source Software Development using AI-Generated Personas

Open source software (OSS) developers often lack direct contact with their users. Issue trackers and forums provide some feedback, but they mostly capture technically savvy users and rarely convey who is affected by a problem or why it matters to them. OSS communities also typically lack the resources to conduct user research, so developers tend to focus on technical problem-solving rather than understanding user needs. Personas can help build empathy, but creating them traditionally requires UX expertise that most OSS projects cannot afford.

We built PersonaFlow, a tool that automatically generates user personas from repository artifacts (e.g., README files, documentation, issue reports) and integrates them into the issue management workflow. Given a repository URL, the tool creates diverse personas, maps them to individual issues with confidence scores, and provides a dashboard that shows which user types are most affected. Developers can edit, merge, or regenerate personas at any point.

We evaluated PersonaFlow in a lab study with 13 OSS developers. Most reported shifts in how they understood their users, and more than half changed their issue responses by adding empathetic language, adjusting explanations for skill level, or raising priority based on user impact. We found two pathways: some developers connected emotionally to personas as real people, while others used them practically for triaging and fairness. Both led to more user-centered behavior. We contribute design implications for AI-assisted persona tools and reframe the lack of empathy in OSS as a structural problem in the information environment rather than an individual failing.