Text Analytics for Predicting Question Acceptance Rates

Online community question answering (CQA) services have gained unprecedented popularity among users wanting to voluntarily exchange solutions without a fee. However, CQA faces two challenges: the growing volume of databases and the increasing number of questions left unanswered. This article proposes classification in text analytics as one way to predict how likely a posted question is to be answered. This involves evaluating the features that characterize the question to understand why community members are or aren´t answering it. Insights from text analytics could help CQA managers guide users regarding posting etiquette, thereby retaining such services´ appeal and ensuring healthy knowledge growth. This study presents a feasible solution to tackle these two problems in CQA, and does so with promising results--particularly in classification by data stream mining with accelerated swarm search feature selection.