Sentiment Analysis based on Hybrid Approach: A Survey

Abstract

Sentiment Analysis is the most widely used text classification technique which is used to analyze any text and result in terms of whether the sentiment of particular text is positive, negative or neutral. Micro blogging is used widely to express opinions like on twitter or Face book with limited character space of 140, one liner is best way to express your opinion. These messages are usually short texts like single sentences and opinions with limited contextual information. Retrieving the sentiment of such short text is a challenging task. It requires a strategy and prior knowledge to extract the correct emotion of post. Like any identification task this will also need a good feature extractor and classification approach. In this paper we are focusing on comparing the feature extractors and also utility of hybrid approach in sentiment analysis.