Insincere Question Classification on Question Answering Forum

Abstract

Insincerity is defined as a word or action that is genuinely not felt by humans, has no meaning, or not based on sincere feelings. In the internet forum, especially question answering forum, insincerity is one of the severe problems because it can affect the quality of internet forums. Frequently found content (i.e., a message or post) which is not appropriate with the rules of the forum in general and can interfere with other users. Machine learning model will be implemented to detect the questions posed by users in the question answering forum. We examine six baseline machine learning models with n-gram, POS tag and word embedding features. The result of our experiment shows that Multilayer Perceptron model using POS tag features give us highest F1-Score, which is 87.81%.