COMPARATIVE ANALYSIS OF TEXT MINING BASED SENTIMENT ANALYSIS MODELS WITH EMOTION DETECTION

Abstract

The Internet has a significant usage of the applications which generate text based data on a regular basis. This data is unstructured which is not directly available in the form of a tabular representation. From the content of emails to the facebook posts, whatsapp messages to feedback messages on e-commerce website, all these text contents form unstructured data on a regular basis. As the data is unstructured, it is indeed a challenging task to analyze it systematically to get meaningful information. Text mining is an emerging field of computer science, which performs analysis of text data in a systematic manner to get insights and to derive some meaningful information. One use of text mining is to detect sentiment or emotion of the writer from his / her text content. This would help in analyzing different aspect of the content writers – from their states of minds to their satisfaction levels. This research work is based on analyzing performances of some most widely used sentiment analysis models with emotion detection.