Question Answering System based on Question Classification and Sentential Level Ranking

and then using the portion of data that is significant for the user's question. For this purpose we need to summarize the data. Summarization is a technique used to extract sentences from a text document that best represent its content. Text summarization (TS) is the process of identifying the most salient information in a document or set of related documents and conveying it in less space (typically by a factor of five to ten) than the original text [1]. There are two types of summarization that is used; content based and context based summarization. The content-based summarization utilizes textual content of the web document while context-based makes use of the hypertext structure of the web and there are basically two known techniques for the summarization. Extractive summaries (extracts) are produced by concatenating several sentences taken exactly as they appear in the main document being summarized. Abstractive summaries (abstracts) are expressed in the words of the summary author. [2] In abstractive summaries the sentences can be rephrased. The proposed implementation uses content based extractive summarization. The techniques like stop word listing and stemming are used to increase the accuracy of the system. The proposed architecture focuses on factoid questions inducing entities and provides the short and relevant answer by using efficient searching and ranking techniques.