Publication Date


Document Type


First Advisor

Pickard, Matthew

Degree Name

B.S. (Bachelor of Science)

Legacy Department

Department of Accountancy


For as long as the stock market, financial news, and financial reports have been around, people have been trying to gain an edge on the market and identify trends in advance, allowing for more intelligent financial decision-making. This has been quite difficult, but in recent years a branch of Natural Language Processing known as sentiment analysis has made this more realistic. By using machine learning or lexicon-based methods to sentiment analysis to analyze the emotion behind text, research has shown potential for investors and companies to identify financial trends in advance that can help them guide strategy or investment decisions. In this project, by analyzing sentiment analysis research over a variety of domains, a lexicon-based approach to sentiment analysis was examined. Its methodologies, benefits, and drawbacks were studied, and through this research, a potential model was proposed that could improve upon the current lexicon-based model and be specifically applied to finance specific works such as financial news articles or financial reports. This paper identifies a potential improved way of doing sentiment analysis for financial news and reports using a lexicon-based approach. By incorporating increased polarity aspects, there is potential to be able to delve to deeper granularity levels when it comes to analyzing sentiment. Typically, a lexicon-based approach to sentiment analysis will depict text as being either positive, negative, or neutral, but the model proposed in this paper describes a way to do this analysis while determining to what extent text is positive or negative. This allows for greater understanding of the sentiment behind financial news and reports and this information can be used to better understand the health of the market, future company earnings potential, a company’s risk levels, and a company’s future ROA numbers among other things.


22 pages




Northern Illinois University

Rights Statement

In Copyright

Rights Statement 2

NIU theses are protected by copyright. They may be viewed from Huskie Commons for any purpose, but reproduction or distribution in any format is prohibited without the written permission of the authors.

Media Type