Survey on Sentiment Analysis of User Reviews

Abstract

User’s sentiments or opinions stated on any website or app have enormous impact on the product sellers, customers, and readers. The formless data from the sites or apps is required to be explored and analyzed. To do this, sentiment analysis has familiar major interest. Sentiment Analysis (SA) is sometimes described as archeology or opinion mining as a research field to analyze people’s feelings or ideas about organizations such as topics, events, individuals, issues, services, products, organizations, and their characteristics. As the number of reviews rapidly increases, online reviews’ sentiment analysis has become a trending topic in research. In this paper, survey is done which highlights the latest research about the implementation of machine learning and deep learning models like support vector machine (SVM), linear regression (LR), Naïve Bayes (NB), convolutional neural networks (CNN), Long short-term memory (LSTM), deep neural networks (DNN), and many others proposed for resolving various challenges in sentiment analysis such as product and app(s) review analysis, textual and visual analysis, classification of sentiments, and cross lingual problems.