Requirements for Automating Moderation in Community Question-Answering Websites

In recent years, community Q&A websites have attracted many users and have become reliable sources among experts from various fields. These platforms have specific rules to maintain their content quality in addition to general user agreements. Due to the vast expanse of these systems in terms of the number of users and posts, manual checking and verification of new contents by the administrators and official moderators are not feasible, and these systems require scalable solutions. In major Q&A networks, the current strategy is to use crowdsourcing with reliance on reporting systems. This strategy has serious problems, including the slow handling of violations, the loss of new and experienced users’ time, the low quality of user reports, and discouraging feedback to new users. While this is a great opportunity to utilize machine-learning approaches to provide automated recommender systems and classification models, there are specific non-functional requirements or aspects related to these software systems that need to be introduced and incorporated in the design of a new system. In this short paper, I pinpoint three key aspects: (1) Any good approach should consider specific attributes and features related to context, (2) Automated mechanisms should be proposed according to the highly evolving content of Q&A websites, and (3) Decisions are best to accompany clear and justifiable explanations. Furthermore, a technical conceptual model is proposed by considering these aspects and related approaches. This is a part of a research project where the final goal is to provide accurate, adaptable, efficient, and explainable solutions for automating moderation actions in Q&A websites.