Predicting Sentiment and Emotion on ASD Using Machine Learning and Deep Neural Networks Based on Twitter Data

Abstract

Autism spectrum disorder (ASD) is by birth neurological disorder. The research on ASD summarizes causes as prenatal conditions, postnatal conditions, and genetic. This paper analyzes the sentiment and emotion of tweets posted in twitter related to ASD causes and treatments. These analytical results provide awareness about treatments and causes of ASD. Natural language processing methods are efficient to perform sentiment and emotion analysis on tweets, but it requires a lot of repeated text processing. So, to reduce the processing time of text, machine learning models are applied on tweets to predict sentiment and emotion. The recent research proved that deep neural networks models are efficient in sentiment analysis and emotion analysis. For efficient results, deep neural networks models are applied on tweets to predict sentiment and emotion. Comparative analysis is performed between ML models results and DNN models results. It is proved that DNN is efficient than ML.