PRODUCT ASPECT RANKING USING DOMAIN DEPENDENT AND DOMAIN INDEPENDENT REVIEW

Abstract

In today’s world, internet is the main source of information. There are many blogs and forum sites available where people discuss on different issues and also almost all ecommerce website provide facility to the users to express opinion about their product and services which is important information available on the internet .The problem with this information is that this reviews are mostly not organized therefore creating difficulty for knowledge acquisition. There are many solution exist to resolve this problem but the available existing methods depends on extracting product aspect only considering single domain relevant review corpus. To address this problem, a method is explored to identify product aspect from online review is by taking into account the difference in aspect statistical characteristic across different corpus. This paper shows need of automatically identifying important product aspects from available online customer review and an approach of aspect ranking. This paper also shows the related work on this domain. Our methodology confirmed product aspect which are less nonspecific in domain independent corpus and more domain specific. Then customer opinion expressed on these aspects is determined using sentiment classifier and finally ranking of product aspect is calculated using it’s ranking relevance score of each aspect .