Discriminative Deep Association Learning based on the Optimized Feature Analysis Adaptive Spider Foraging Model for twitter sentiment analysis

Abstract

The sentiment aspect is an important discussion in social media to discuss various forums like the product, social events, monuments, etc. All over, social media is the most dominant factor in discussion forums user comments as tweets be problematic analyses. Due to increasing sarcasm in social media terms contain sentiment terms and behaviors of users, the importance of features in data analyses needs more deep evaluation to improve the accuracy. To propose a Discriminative Deep Association Learning based on the Optimized Feature analysis Adaptive Spider Foraging Model (ASFM) to predict the occurrence of the event in social media terms. The method utilizes the tweets and messages generated from a social network with Tweet term features. Initially, the progress begins with the preprocessing of the social media terms and Tweet term facts to identify the features. Because of the sentimental side of sarcasm, the Semantic Entropy Vector Transformation model detects both sarcasm and non-sarcasm weights as features. Social foraging models identify optimal features based on fitness weights. The tweets and the structure of tweet words are analyzed and grouped into classes based on a semantic ontology process.