|Project Description: ||Question Answering System (QAS) is a specialized area in the field of information retrieval, whose main objective is to obtain precise answers to a user’s questions. Most of the QAS deal with morphological and syntactic levels of processing. They concentrate on words or phrases or syntactic substructures of the sentences. Problems arise when there is a question such as “who is the outstanding player in 2010?” and an answer sentence of “Nadal is the prominent player in 2010” in the document set. A question answering system cannot retrieve the answer with only morphological and syntactic processing. Semantic level processing is required to match “outstanding” with “prominent”.
We present a Semantic QAS which deals with answering “Who is” questions using semantic tag clouds. The major goal of this project is to identify the answer sentences which contain semantically related question keywords and show the important role of semantic processing. Input of this project is a “who is” question and a group of documents. The system will process the documents, question and finds the main and possible answers to the question. The output is the answer for the “who is” question. The evaluated system produces answers to a set of “who is” questions, even when the question and answer sentences have different but semantically related keywords.