Procaci, Thiago Baesso, et al. "How to find people who can help to answer a question?–Analyses of metrics and machine learning in online communities." Computers in Human Behavior 51 (2015): 664-673.

This study focuses on enhancing the responsiveness and reliability of online communities by identifying users who consistently provide helpful and accurate responses, referred to as reliable users. Through examining various attributes of online communities and their users, the research seeks to find a correlation between these characteristics and the users' reputation. Two machine learning techniques, artificial neural networks (ANN) and clustering algorithms, are proposed to analyze user attributes for identifying reliable sources. The findings indicate that ANNs are highly effective, with approximately 90% accuracy in identifying reliable users, while clustering algorithms facilitate the grouping of these users, thereby improving the overall quality of interactions within online communities.