Comparison between Machine Learning Algorithms Used for Sentiment Analysis
International Journal of Advanced Research in Engineering and Technology (IJARET)
9 Pages Posted: 25 Jul 2021
Date Written: December 2020
Sentiment analysis otherwise called opinion mining is the artificial intelligence technique used to analyze and predict the user emotions which can be positive, negative or neutral in a given text using some of the machine learning & deep-learning libraries or models. It is one of the most trending ongoing research area in the artificial intelligence, most of the researchers and business models use this approach for the text mining and analyzing the human sentiments. It is very dif icult to analyze a large number of data set manually this is where the sentiment analysis model comes into play, movie reviews can be used as dataset for training and preprocessing the data. Sentiment analysis on the raw text is a very complicated task due to various reasons such as a sarcastic text or positive and negative sentiment used in the same text. using machine learning algorithms such as logistic regression and naive bayes classification over a cleaned and processed data which is obtained after using tf-idf vectorization on the dataset. This model gives the accuracy and precision of around 95% and can be used in various business models.
Keywords: Machine Learning, Natural Language Processing (NLP), supervised machine learning algorithm, logistic regression, naive bayes classification, tf-idf vectorization
Suggested Citation: Suggested Citation