Mining topical relations between opinion word and opinion target

Abstract

Opinions are plays very important role in today's life for purchasing any item or product. This opinion contains lots of information which can be useful for future systems. Opinion mining is an emerging field which combines lots of tasks to find the correct Opinion Words and Targets which describes the product or express the sentiment towards product. In this paper our main focus is on Topical Relation, which shows the relation between opinion target and opinion words. These topical relations are useful to gain the idea of which topics are mostly discussed by customers for particular product. We will also calculate the sentiment of customer reviews which classifies into three categories positive, negative and neutral.