Sentiment Analysis: A General Review and Comparison

The use of natural language processing and opinion mining methods has been utilized throughout the last couple of years through sentiment analysis to detect, obtain, compute, and examine information which could be valuable to users. Organizations that make use of these methods need these methods to evaluate and improve their customer feedback. There are different types of techniques that have been previously implemented. Sentiment analysis tools can be categorized as machine learning techniques and lexicon-based techniques. The machine learning techniques are divided into supervised and unsupervised learning while lexicon-based techniques are split into dictionary-based and corpus-based approaches. The machine learning techniques categorize the polarity in sentiments while the lexicon-based techniques utilize sentiment lexicons. In this work, different machine learning and lexicon-based techniques have been reviewed to discuss their advantages and their limitations when implemented.