A semantic approach for implicit topical term identification and ranking

Opinion mining is the process of retrieving the opinion evidences from the reviews and finding their polarity such as positive, negative. Topical term mining enhances the opinion mining in which topical terms have to be extracted along with opinion evidences. Topical terms are specific attribute or quality of a topic. This paper proposes an approach for topical term mining. It aims to assist the users in decision making utilizing the score table. Score table comprises the topical terms of a specific topic along with its positive, negative scores and neutral counts individually. Proposed work incorporates the ontology for identifying implicit topical terms and the variation of maximum opinion score algorithm for finding the polarity. Goals of this approach are 1) to improve the identification of implicit topical terms using ontology 2) to provide accurate estimation of sentiment polarity.