Detecting Interpersonal Conflict in Issues and Code Review: Cross Pollinating Open- and Closed-Source Approaches

Huilian Sophie Qiu
Bogdan Vasilescu
Christian Kästner
Emerson Rex Murphy-Hill
International Conference on Software Engineering: Software Engineering on Society(2022)


In software engineering, interpersonal conflict in code review, such as toxic language or an unnecessary pushback on a change request, is a well-known and extensively studied problem because it is associated with negative outcomes, such as stress and turnover. One effective approach to prevent and mitigate toxic language is to develop automatic detection. Two most-recent attempts on automatic detection were developed under different settings: a toxicity detector using text analytics for open source issue discussions and a pushback detector using logs-based metrics for corporate code reviews. While these settings are arguably distinct, the behaviors that they can capture share similarities. Our work studies how the toxicity detector and the pushback detector can be generalized beyond the respective contexts in which they were developed and how the combination of the two can improve interpersonal conflict detection. This research has implications for designing interventions and offers an opportunity to apply a technique to both open and closed source software, possibly benefiting from synergies, a rarity in software engineering research, in our experience.

Research Areas