Advanced technology development and high speed internet connection have connected people across the world. Every business marks its presence on web and directly gets connected with the interested customers. On the other hand, consumers express their opinion about the services and products on companies’ website, forum or any other social platform. Therefore, as data is available in bulk, it is very difficult for manufacturers and buyers to make decision due to vagueness. This paper proposes an enhanced opinion mining approach that filters ambiguity and highlights general sentiments of a particular product. This aspect based opinion mining approach has three important phases, prepossessing followed by feature extraction and then ranking.

Keywords : Sentiment Analysis; Opinion Mining; Polarity; Machine Learning.