Does Applying Deep Learning in Financial Sentiment Analysis Lead to Better Classification Performance?

Authors

Tao Wang, CUNY Queens CollegeFollow
Changhe Yuan, CUNY Queens CollegeFollow
Cuiyuan Wang, CUNY Graduate Center

Document Type

Article

Publication Date

4-2020

Abstract

Using a unique data set from Seeking Alpha, we compare the deep learning approach with traditional machine learning approaches in classifying financial text. We apply the long short-term memory (LSTM) as the deep learning method and Naive Bayes, SVM, Logistic Regression, XGBoost as the traditional machine learning approaches. The results suggest that the LSTM model outperforms the conventional machine learning methods on all metrics. Based on the tSNE graph, the success of the LSTM model is partially explained as the high-accuracy LSTM model distinguishes between positive and negative important sentiment words while those words are chosen based on SHAP values and also appear in the widely used financial word dictionary, the Loughran-McDonald Dictionary (2011).

Comments

Originally published as: Wang, Cuiyuan, Tao Wang, and Changhe Yuan. ''Does Applying Deep Learning in Financial Sentiment Analysis Lead to Better Classification Performance?'' Economics Bulletin, Vol. 40, No. 2, 2020, pp. 1091-1105.