Opinion Mining Using Domain Relevance

Opinion mining is determining people¿s opinions, about entities such as products, services and their attributes. Opinion feature identification from online reviews evaluated in two domain corpora, one is domain specific and other is domain independent corpus and this evaluation is based on number of occurrence of that feature. Domain relevance is used to measure distribution. By applying a set of syntactic rules identify candidate features in user reviews is primary task. Features extracted from this are specific to a domain. For each extracted candidate feature extrinsic domain relevance and intrinsic domain relevance value are calculated. Experimental result show that extracted opinion feature remains constant after specific threshold value. Precision value decreases according to the threshold value increases.