A survey on textual entailment based question answering

Abstract

Question answering, an information retrieval system that seeks knowledge, is one of the classic applications in Natural Language Processing. A question answering system comprises numerous sets of subtasks. Some of the subtasks are Passage Retrieval, Answer Ranking, Question Similarity, Question Generation, Question Classification, Answer Selection, and Answer Validation. Numerous approaches have been experimented on in the question answering system to achieve accurate results. One such approach for the question answering system is Textual Entailment. Textual Entailment is a framework that captures significant semantic inference. Textual Entailment of two text fragments can be defined as the task of deciding whether the meaning of one text fragment can be inferred from another text fragment. This survey discusses how and why Textual Entailment is applied to various subtasks in question answering.

Keywords

• Natural Language Processing
• Question Answering
• Textual Entailment