Question Answering via Semantic Web Service Composition

Web question answering (WQA) and Semantic Web services are currently two separate fields in intelligent web research, but in practical application, they are usually highly correlative. This article combines the techniques in Natural Language Processing (NLP) and Hierarchical Task Network (HTN), and puts forward a WQA framework that is based on Semantic Web service composition. This framework uses NLP to analyze questions, describes the questions with predicate logic, maps them to relative Semantic Web services, and transforms question answering to a planning problem by constructing question domain and planning domain. It dynamically composites relative services making use of HTN system SHOP2, gets the planning series with Semantic Web services, and finally obtains the result that satisfied customers’ need.