Apply FNN Model to Construct Ontology-based Q&A System

The Q&A (Question & Answer) knowledge base of an organization is important for knowledge management. It can provide experience and knowledge for a questioner. This paper applies a Fuzzy-Neural Network (FNN) model to construct Q&A knowledge base automatically. The FNN computes the important degree with sentences and infers the strength of each sentence for question properties. In addition, the back-propagation learning algorithm is adopted to train the question extractor. Moreover, we utilize natural language processing technology to analyze sentences of Chinese documents, and make use of Chinese knowledge bases to aid the selection of questions. The experimental results exhibit that the proposed approach can extract questions from Chinese documents effectively. Key-Words: Ontology, Knowledge Base, Fuzzy Inference, Neural Network, Q&A, Knowledge Management.