BERT for stock market sentiment analysis
Publikation: Beiträge in Sammelwerken › Aufsätze in Konferenzbänden › Forschung › begutachtet
Standard
IEEE 31st International Conference on Tools with Artificial Intelligence: ICTAI 2019 : proceedings : 4-6 November 2019, Portland, Oregon. Piscataway: IEEE - Institute of Electrical and Electronics Engineers Inc., 2019. S. 1589-1593 ( International Conference on Tools with Artificial Intelligence, ICTAI).
Publikation: Beiträge in Sammelwerken › Aufsätze in Konferenzbänden › Forschung › begutachtet
Harvard
APA
Vancouver
Bibtex
}
RIS
TY - CHAP
T1 - BERT for stock market sentiment analysis
AU - Sousa, Matheus Gomes
AU - Sakiyama, Kenzo
AU - Rodrigues, Lucas De Souza
AU - Moraes, Pedro Henrique
AU - Fernandes, Eraldo
AU - Matsubara, Edson Takashi
N1 - Conference code: 13
PY - 2019/11
Y1 - 2019/11
N2 - When breaking news occurs, stock quotes can change abruptly in a matter of seconds. The human analysis of breaking news can take several minutes, and investors in the financial markets need to make quick decisions. Such challenging scenarios require faster ways to support investors. In this work, we propose the use of bidirectional encoder representations from transformers BERT to perform sentiment analysis of news articles and provide relevant information for decision making in the stock market. This model is pre-trained on a large amount of general-domain documents by means of a self-learning task. To fine-tune this powerful model on sentiment analysis for the stock market, we manually labeled stock news articles as positive, neutral or negative. This dataset is freely available and amounts to 582 documents from several financial news sources. We fine-tune a BERT model on this dataset and achieve 72.5% of F-score. Then, we perform some experiments highlighting how the output of the obtained model can provide valuable information to predict the subsequent movements of the Dow Jones Industrial (DJI) Index.
AB - When breaking news occurs, stock quotes can change abruptly in a matter of seconds. The human analysis of breaking news can take several minutes, and investors in the financial markets need to make quick decisions. Such challenging scenarios require faster ways to support investors. In this work, we propose the use of bidirectional encoder representations from transformers BERT to perform sentiment analysis of news articles and provide relevant information for decision making in the stock market. This model is pre-trained on a large amount of general-domain documents by means of a self-learning task. To fine-tune this powerful model on sentiment analysis for the stock market, we manually labeled stock news articles as positive, neutral or negative. This dataset is freely available and amounts to 582 documents from several financial news sources. We fine-tune a BERT model on this dataset and achieve 72.5% of F-score. Then, we perform some experiments highlighting how the output of the obtained model can provide valuable information to predict the subsequent movements of the Dow Jones Industrial (DJI) Index.
KW - BERT
KW - NLP
KW - Sentiment analysis
KW - Stock market
KW - Informatics
KW - Business informatics
UR - http://www.scopus.com/inward/record.url?scp=85081094676&partnerID=8YFLogxK
U2 - 10.1109/ICTAI.2019.00231
DO - 10.1109/ICTAI.2019.00231
M3 - Article in conference proceedings
AN - SCOPUS:85081094676
SN - 978-1-7281-3799-5
T3 - International Conference on Tools with Artificial Intelligence, ICTAI
SP - 1589
EP - 1593
BT - IEEE 31st International Conference on Tools with Artificial Intelligence
PB - IEEE - Institute of Electrical and Electronics Engineers Inc.
CY - Piscataway
T2 - IEEE International Conference on Tools with Artificial Intelligence
Y2 - 4 November 2019 through 6 November 2019
ER -