Sentiment Dictionary for Effective Detection of Web Users' Opinion

In this paper we present a sentiment dictionary that aids in identifying subjective phrases from opinionated texts useful for opinion detection approaches in determining opinion of web users from opinionated texts. Web users document their opinion on diverse topics using variety of mediums available in the web. These opinions will be very useful to other web users that assist them to change their perspective or take a few useful decisions. Today, opinions are expressed by web users using multimodal opinion elements like normal phrases, emoticons and short words or sms language. To capture web users opinion with multimodal opinion elements, we need dictionaries that will assist in identifying these multimodal opinion phrases. Dictionaries mentioned in literatures are less useful in identifying multi modal subjective phrases and opinion of web users. We examine a number of dictionaries and propose an effective sentiment dictionary useful in identifying subjective phrases and opinion of web users on products. The result obtained indicates the proposed sentiment dictionary is effective in mining multimodal user opinion.