AskDragon: a redundancy-based factoid question answering system with lightweight local context analysis

ABSTRACT

We introduce our QA system AskDragon which employs a novel lightweight local context analysis technique to handling two broad classes of factoid questions, entity and numeric questions. The local context analysis module dramatically improves the efficiency of QA systems without sacrificing high accuracy performance.

INDEX TERMS

Primary Classification:
  H. Information Systems
  H.3 INFORMATION STORAGE AND RETRIEVAL
      H.3.4 Systems and Software
          Subjects: Question-answering (fact retrieval) systems**

General Terms:
Algorithms, Experimentation, Performance

Keywords:
answer generation, answer scoring, local context analysis, question answering, redundancy-based approach