A Survey on Sentiment Analysis
From defining the sentiment analysis to algorithms for sentiment analysis and from the first step of sentiment analysis to evaluating the predictions of sentiment classifiers, additional feature extractions to boost performance are discussed with practical results.
Abstract
Sentiment analysis is one of the recent technologies under NLP (an application of Al and ML). Sentiment analysis is used in many applications for recommendation and feedback analysis. In this paper, from defining the sentiment analysis to algorithms for sentiment analysis and from the first step of sentiment analysis to evaluating the predictions of sentiment classifiers, additional feature extractions to boost performance are discussed with practical results. A brief description of complex sequence-based Neural Network sentiment classifiers with reasonable analytics is provided. The practical results declared in this paper are from the implantation of sentiment analysis on the IMDB movie reviews dataset. Evaluation metrics such as accuracy, precision, recall, and f1-score are used. This Research-based survey has been divided into different sections, each section concerning the stepwise process of sentiment analysis.