Systematic Comparison of Question Target Classification Taxonomies Towards Question Answering.

Question target classification is one of the essential research topics in question answering. Accurate identification and classification of question targets can help understand questions for retrieving relevant passages and assist answer extraction and ranking for improving answer retrieval accuracy and user satisfaction with return answers. This paper presents a systematic analysis on question target classification taxonomy. We investigate existing definitions of the classification and propose a concise definition. We then compare the existing classification taxonomies. The relevancies of the taxonomies are analyzed, inspiring us to propose a new taxonomy classification strategy. We finally summarize the characteristics and tendency of the current research of question target classification taxonomy. The systematic comparison is expected to provide consistent and meaningful guidance in the research of question target understanding in question answering.