SENTIMENT ANALYSIS AND SARCASM DETECTION

Abstract

The study of sentiment analysis for this work began by going through the present research conducted by different researchers. While performing the thorough literature review, comparative and qualitative analysis of different research papers were performed keeping in mind the datasets used by various researchers, the method used to derive the sentiment, categories of the outcome, areas in which sentiment analysis and others was studied. The summary of the entire study is culminated in a tabular format in the upcoming chapters. After going through the literature, it was found that there is a huge area for studying and understanding sarcasm which is on the textual format. Also, since, detecting sarcasms can further pave way for understanding the hidden sentiment or the true sentiment, Sarcasm detection was further studied. In the study for detecting sarcasm, we design a model that takes care of the fine sentiments and the coarse sentiments in the comment / statements under the study. This model is also tested on state of art models and also on 3 different datasets and it gives good results.