Sentiment analysis of financial news articles using performance indicators

Abstract

Mining financial text documents and understanding the sentiments of individualinvestors, institutions and markets is an important and challenging problem in the literature.Current approaches to mine sentiments from financial texts largely rely on domain-specificdictionaries. However, dictionary-based methods often fail to accurately predict the polarityof financial texts. This paper aims to improve the state-of-the-art and introduces a novel sentimentanalysis approach that employs the concept of financial and non-financial performanceindicators. It presents an association rulemining-based hierarchical sentiment classifiermodelto predict the polarity of financial texts as positive, neutral or negative. The performance ofthe proposed model is evaluated on a benchmark financial dataset. The model is also comparedagainst other state-of-the-art dictionary and machine learning-based approaches andthe results are found to be quite promising. The novel use of performance indicators forfinancial sentiment analysis offers interesting and useful insights.