Enhanced Artificial Bee Colony Enabled Neutrosophic Clustering for Predicting Polarization of Education Data Sentiment Analysis

Abstract

The Sentiment Analysis is the process of identifying and classifying public’s opinion from a part of text into sentiment which comprised of positive, negative or neutral attitudes related to the specific object or subject. In the area of Nature language processing, sentiment analysis plays a vital role for education field, as the feedback from students and parents is very essential to evaluate the efficacy of learning knowledges. The review about the course content and education quality by collecting education tweets may greatly influence to improve the quality of teaching by providing positive and negative sentiments. This paper focuses on developing an unsupervised model for predicting education sentiment analysis using the tweet dataset. The uncertainty and indeterminacy in categorizing the tweets as positive or negative, is well handled by applying Neutrosophic C Means Clustering (NCMC), where it represents each instances in terms of membership towards truth value, falsity value and indeterminacy. The clustering process itself is enhanced by adapting Artificial Bee Colony Algorithm (ABC), which selects the centroids using their food source searching behaviour. The performance of Artificial Bee Colony Algorithm enabled Neutrosophic C Means Clustering (ABC-NCMC) for predicting education sentimental analysis is compared with DBSCAN and Fuzzy C Means. The results proved that ABC-NCMC achieves highest rate of accuracy with less error rate for predicting education tweet sentiment analysis