Literature Review on Sentiment Analysis in Social Media: Open Challenges toward Applications

Abstract

In today’s world, sentiment analysis stands as the prominent research topic in demand under the Natural Language Processing (NLP). The fundamental objective of this research topic is to spot out the emotions and opinions of the customers or users via a text basis. Even though numerous research works have been carried put in this field through diverse models, sentiment analysis is still considered a challenging problem with so many conflicts to be solved. Some of the existing challenges are due to the slang words, new accents, grammatical and spelling mistakes etc. This paper plans to make a literature review using different machine learning algorithms with various data. The current literature review temps to survey nearly 20 contributions, which covers different types of applications being used for sentimental analysis. At first, the analysis focuses on illustrating the contributions of each work and observes the type of machine learning algorithms used. Moreover, the analysis also concentrates on the identification of the type of data used. Further, the utilized environment and the performance measures covered in each work is evaluated, and concluded with proper research gaps and challenges, which helps to identify the non-saturated application for which the sentimental analysis is needed most in upcoming research.