InsunTourQA: A Restricted-Domain Question Answering system

Restricted-Domain Question Answering (RDQA) system works on specific domains and often uses document collections restricted in subject and volume. This paper presents the design and implementation of a RDQA system: InsunTourQA, which is face to the domain of Chinese tourism. Natural language processing technologies, including Chinese word segmentation, Part of Speech tagging, named entity recognition and chunking, are involved in this system. The whole framework consists of 5 subsystems: Query processing, information retrieval, answer extraction, web information processing and knowledge management. Text Classification method is used to reduce the irrelevance candidates and the system reaction time. A cover-based technology is used in the information retrieval. A new semantic similarity calculating method is proposed to measure the similarity between sentences. Experiments show the top10 MRR of our system is nearly 60%.