From Questions to Insights: Exploring XAI Challenges Reported on Stack Overflow Questions

The lack of interpretability is a major barrier that limits the practical usage of artificial intelligence (AI) models. Several eXplainable AI (XAI) techniques (e.g., SHAP, LIME) have been employed to interpret these models' performance. However, users often face challenges when leveraging these techniques in real-world scenarios. Therefore, they submit questions in technical Q\&A forums like Stack Overflow (SO) to resolve these challenges. To support explainability, we conduct an exploratory study to expose these challenges, their severity, and features that can make XAI techniques more accessible and easier to use. Our contributions to this study are fourfold. First, we manually analyzed 663 SO questions that discussed challenges related to XAI techniques. Our careful investigation produced a catalog of seven challenges (e.g., disagreement issues). To further validate, we collected feedback from 52 practitioners, who agreed that they encountered these challenges and reported no additional ones. Our analysis and practitioners' feedback showed that model integration issues and disagreement issues emerged as the most prevalent challenges. Second, we attempt to estimate the severity of each XAI challenge by determining the correlation between challenge types and answer metadata, such as receiving answers, including the acceptable ones. Our analysis suggests that model integration issues is the most severe challenge. Third, we attempt to perceive the severity of these challenges based on practitioners' ability to use XAI techniques effectively in their work. Practitioners' responses suggest that disagreement issues most severely affect the use of XAI techniques. Fourth, we seek input from practitioners on improvements or features that could make XAI techniques more accessible and user-friendly. The majority of them suggest consistency in explanations and simplified integration. Our study findings might (a) help to enhance the accessibility and usability of XAI and (b) act as the initial benchmark that can inspire future research.