A Survey of Sentiment Analysis Using Various Machine Learning and Deep Learning Techniques

Abstract

In the modern era, Deep Learning (DL) and Machine Learning (ML) emerged as the advanced strategies for applying various Natural Language Processing (NLP) to enhance the performance of sentiment analysis. Main of this sentiment analysis is to extract the essential   opinions as well as attitudes towards an entity.  This process is one of the most powerful tool utilized by marketing, governments, business and others. Also, traditional sentiment analysis approaches are mainly focused on text content. Nevertheless, advance technology have permitted to prompt their feelings and expressions through video, audio, texts and images. As the most crucial task for improving decision-making, sentiment analysis involves identifying the underlying sentiment or opinion of data. Though sentiment analysis has advanced recently, there are still issues with current models, including excessive dimensionality, negation handling, domain dependence, and ineffective keyword extraction. This study looks at several viewpoints about the development and application of a successful sentiment analysis model such as sequential ML and DL techniques and conducts the detailed review.