Deep Learning Approaches for Sentiment Analysis Challenges and Future Issues

Abstract

A lot of user-generated data is available these days from huge platforms, blogs, websites, and other review sites. These data are usually unstructured. Analyzing sentiments from these data automatically is considered an important challenge. Several machine learning algorithms are implemented to check the opinions from large data sets. A lot of research has been undergone in understanding machine learning approaches to analyze sentiments. Machine learning mainly depends on the data required for model building, and hence, suitable feature exactions techniques also need to be carried. In this chapter, several deep learning approaches, its challenges, and future issues will be addressed. Deep learning techniques are considered important in predicting the sentiments of users. This chapter aims to analyze the deep-learning techniques for predicting sentiments and understanding the importance of several approaches for mining opinions and determining sentiment polarity.