This paper is published in Volume-8, Issue-4, 2022
Area
Computer Science
Author
Sagar Bansal, Snigdha Patil
Org/Univ
Independent Researcher, India
Pub. Date
25 July, 2022
Paper ID
V8I4-1201
Publisher
Keywords
BERT Model, Stock Market Direction, Natural Language Processing, Sentiment Analysis

Citationsacebook

IEEE
Sagar Bansal, Snigdha Patil. Predicting stock market direction: hit songs’ sentiment analysis by using bert, International Journal of Advance Research, Ideas and Innovations in Technology, www.IJARIIT.com.

APA
Sagar Bansal, Snigdha Patil (2022). Predicting stock market direction: hit songs’ sentiment analysis by using bert. International Journal of Advance Research, Ideas and Innovations in Technology, 8(4) www.IJARIIT.com.

MLA
Sagar Bansal, Snigdha Patil. "Predicting stock market direction: hit songs’ sentiment analysis by using bert." International Journal of Advance Research, Ideas and Innovations in Technology 8.4 (2022). www.IJARIIT.com.

Abstract

Various machine learning algorithms have been used to predict stock market prices and perform analyses. Investor sentiment has been taken into account to predict the prices. However, there are different ways in which investor mood or sentiment can be influenced. Studies have confirmed that people’s mood or sentiment influences their choice of songs and likewise, songs influence people’s mood or sentiment and in fact their buying behavior. This paper proposes a prediction of whether the stock market will be bearish or bullish on the basis of daily hit songs listened to by people on a popular music streaming platform. Only a few types of research have been conducted that propose a correlation between popular songs’ sentiments and predicting the stock market. In this paper, the direction of the Dow Jones Industrial Average (DJIA) index is predicted from the lyrics of daily hit songs on Spotify listened to in the region of the United States of America. The model is trained on the dataset of lyrics of daily top 50 songs on Spotify from January 2017 to February 2022. The BERT (Bidirectional Encoder Representations from Transformers) model in Natural Language Processing (NLP) has been used to predict the direction of the DJIA index for the next day.