Question Answering (QA) is a narrowed form of information retrieval which filer the information from pool of documents as per the users demands. Question Answering systems uses natural language processing techniques (NLP) to process a question and then searches for the needed information to identify the answer and presents the answer to the user. With increasing amounts of data it becomes more and more difficult for users to derive material of interest, to search efficiently for specific content or to gain an overview of influential, important and relevant material so to solve this problem information filtering plays an important role to reduce data and provide user a precise set of data instead of overloading users with a large number of irrelevant documents. Requesters visiting a CQA site need to have a sufficient level of confidence that their questions will be answered, in order to come to the site in the first place. Consequently, it is critical to keep the rate of unanswered questions to a minimum. This paper aims to help alleviating this problem by using an automatic question answering system using cosine similarity and ranking algorithm which gives output in minimal time.