Exploitation of Newspaper-article Characteristics for Article Retrieval and Answer Extraction in QAC Task 2
暂无分享,去创建一个
In this paper, we discuss a system for newspaper article retrieval and answer extraction. Due to the rapidly increasing amount of accessible information, systems that allow search in natural language are expected to play a much more important role in the very near future. Our system, called RAIK-Prassie, is designed for TASK 2 of QAC. The design of the RAIK-Prassie system focuses mainly on practical use of characteristics of newspaper articles and on questions related to person names. In addition, we implement the system such that it replies at most one answer to each query.
[1] Sadao Kurohashi,et al. Dialogue Helpsystem based on Flexible Matching of User Query with Natural Language Knowledge Base , 2000, SIGDIAL Workshop.