Complex Cross-lingual Question Answering as a Sequential Classification and Multi-Document Summarization Task
Hideki Shima, Ni Lao, Eric Nyberg, Teruko Mitamura · 2008
In this paper, we describe the JAVELIN IV system, which treats complex question answering as a sequential classification and multi-document summarization task. Our research and development effort is based on various forms of linguistic annotation, and a comparison of various answer extraction and summarization algorithms. We discuss the use of different units of extraction, the effect of different syntactic features for classification, and the effect of different summarization strategies. We also analyze how the performance of machine translation and information retrieval affect the performance of question answering. In the NTCIR-7 CCLQA main track official evaluation, our system achieved 16.3 % and 19.2 % accuracy in the English-to-Japanese and English-to-Chinese subtasks, respectively.