A Survey on Frameworks and Methods of Question Answering

Question Answering provides a new manner for information query. Compared to existed search engines, Q&A systems are able to return answer for the questions represented by natural language. Hence, Q&A is always one of core reseraches in the field of information processing. In this paper, we present survey targets the state of the art in Q&A under systems and approaches. Firstly, we introduce the basic architecture of a typical Q&A system. Then, we present ten typical Q&A systems in different views including developer, supporting language, information source, answer type, etc. In detail, we investigate the key research endeavors on Q&A. We summary the exploited approaches to questions classification, questions similarity, answer extraction, answer sorting and question translation. Finally, we will also present our perspective on future directions of Q&A.