Oil Sector and Sentiment Analysis—A Review

Abstract

Oil markets reveal considerably volatile behaviour due to a range of factors. Exogenous factors, such as the COVID-19 pandemic and ongoing wars and conflicts, impose even more difficulties for prediction purposes. As a tool to better understand and improve forecasting models, many researchers are using sentiment analysis techniques to identify the sentiments being emanated in the news and on social media. Following the PRISMA standards, this work systematically reviewed 34 studies out of 320 from the Scopus and Web of Science databases. The results indicate that one can use several different sources to construct a text dataset and develop a sentiment analysis. For instance, Reuters, Oilprice.com, and Twitter are among the more popular ones. Among the approaches used for extracting public sentiment, it became apparent that machine learning-based methods have been increasing in prevalence in recent years, both when applied alone and in conjunction with lexicon-based methods. Finally, regarding the purpose of employing sentiment analysis, the most favourable goal for collecting sentiments concerning the oil market is to forecast oil prices. There is a consensus among the authors that sentiment analysis improves the quality of predictive models, making them more accurate. This work aims to assist academics, researchers, and investors interested in the oil sector.