A systematic review of NLP methods for Sentiment classification of Online News Articles

Natural language processing (NLP) has developed into an essential field that includes sentiment analysis of news articles. It consists of the automatic identification of sentiment in articles using NLP models, which might give businesses, decision-makers, and the general public information that is helpful. Many NLP models have been published in this field for the sentiment analysis of online news articles. Some of the models in this area include VADER, TextBlob, Support Vector Machines (SVM) and Naive Bayes, Recurrent Neural Networks (RNNs), and Transformer-based models. The rule-based technology VADER provides sentiment scores to the text’s words. TextBlob classifies a text’s sentiment using machine learning methods. The methods for probabilistic machine learning Text are categorized using Naive Bayes and SVM based on the frequency of phrases. RNNs analyze text one word at a time and use the previous words to predict how the next word will sound. Transformer-based models can be customized for sentiment analysis and come pre-trained on a lot of text, including BERT, RoBERTa, and GPT-2. The model selection depends on the project’s unique requirements. Overall, NLP models for sentiment analysis of online news articles have the potential to offer insightful information for decision-making across a range of fields. In this paper, the review of different approaches and models for performing sentiment classification on online news articles is provided and the reader may be able to find the application of sentiment analysis in the domains of their interest.