New Lexicon for Financial Sentiment Analysis

What is it about?

Understanding how people feel about money and the stock market is increasingly important. Financial Sentiment Analysis (FSA) is a powerful tool that helps us figure out what investors think and predict how the financial markets might behave. But finance is a field that uses a lot of specialized language, for example, words like "liability" and "debt" might sound very bad in everyday talk, but they're not always negative, as in "solved_debt". To tackle this challenge, we've come up with something called FinSenticNet. Constructing FinSenticNet involves several steps, like breaking down and extracting financial concepts, creating a graph, and inferring the sentiment for all concepts. Our tests show that FinSenticNet does a really good job of understanding sentiments, with high accuracy and good scores, when we tested it on different datasets.

Featured Image

Photo by GuerrillaBuzz on Unsplash

Photo by GuerrillaBuzz on Unsplash

Why is it important?

Lexicon-based sentiment analysis has unique advantages, such as being highly interoperable, more reproducible, and better explainable. Among various other financial sentiment lexicons, FinSenticNet stands out due to its ability to capture domain-specific language features and at the same time to cover broader phrases. These features improve dealing with jargon, terminologies, and collocations in finance language.