Optimization of Existing Sarcasm Detection Models using Fuzzy Neutrosophic Techniques

Abstract

Neutrosophy, defined as the lack of bias of a word in the semantic representation of textual information, is a scientific novelty that is utilized to numerically speak to the set of three of ideas - truth, indeterminacy and lie. Notwithstanding, certifiable content can't cautiously be doled out into one of the three classes as it relies upon the point of view, feelings, and comprehension of both the peruser and the author. Therefore, conversion of the basic concept of Neutrosophy into a Polar Fuzzy Neutrosophic technique and applying semantic nets to applications of sentiment analysis proves to improve the accuracy and improvement of existing models. The objective of our research is, therefore, to introduce the idea of Neutrosophy and build a polar semantic chart from a data set to speak to the extent of the extremity of opinions for various components of the content and use the technique described to optimize existing sarcasm detection models. The segregation of texts into the three sets of ideas and the subsequent creation and extraction of features in existing models will result in a more accurate solution to the problem.