Using weighted features to predict questions’ answerability in question and answer communities

What is it about?

This study developed WANQA, a tool designed to predict whether a question on online Q&A platforms is likely to be answered when it's first submitted. By providing real-time feedback, WANQA helps users refine their questions, improving clarity and increasing the chance of receiving an answer. WANQA uses 20 common features from most Q&A platforms, focusing only on what is available at the time of submission, making it adaptable to different platforms. After testing several methods, the researchers found that a Naive Bayes algorithm with 50% of selected features performed best in terms of speed and accuracy. Unlike previous models that depend on platform-specific features like votes or user reputation, WANQA works across various platforms, improving user experience and content quality. This approach promotes better knowledge-sharing, helping diverse users engage more effectively in Q&A communities.

Featured Image