Information Extraction from WWW using Structural Approach

The expeditiously growing World Wide Web (WWW) consists of a large number of data sources from diverse organizations all over the world. The heterogeneous nature of these data poses challenges to researchers to extract specific information. In this regard, the area of finding answers of a specific question from available web contents is an emerging area of research. Questions are normally expressed in natural language and for finding answers to natural language questions from web contents; Question Answering (QA) is the most promising framework, which can be implemented on either closed domain or open domain. In this paper, we propose an automated QA system which can answer binary and wh-interrogated questions about closed domain using wikipedia articles as its knowledge source. The system allows us to generate questions from wikipedia pages and then to extract answers to questions from wikipedia pages in real time.