TY - JOUR
T1 - Leveraging Stock Discussion Forum Posts for Stock Price Predictions
T2 - Focusing on the Secondary Battery Sector
AU - You, Jisoo
AU - Jang, Haryeom
AU - Kang, Minsuk
AU - Yang, Sung Byung
AU - Yoon, Sang Hyeak
N1 - Publisher Copyright:
© 2013 IEEE.
PY - 2024
Y1 - 2024
N2 - The advent of social media has transformed the stock investment landscape considerably. Individual investors are increasingly shifting from traditional financial institutions to platforms such as Facebook, X, and stock discussion forums to obtain information and exchange opinions, significantly affecting the stock market. Post-COVID-19, the influx of individual investors has accentuated their importance in the South Korean stock market, necessitating the development of prediction models utilizing data from stock discussion forums, a facet of social media. Research on predicting stock price fluctuations using social media data is lacking, and most studies have focused on news articles, with insufficient utilization of stock discussion forum posts. Considering the significant proportion of individual investors in the Korean stock market, it is imperative to conduct research utilizing posts from stock discussion forums, where individual investors' opinions are directly formed, for accurate stock price predictions. This study specifically focuses on the secondary battery sector, which garnered significant attention during the pandemic, to examine the sentiments of individual investors. Employing data from stock price prediction models, this study compares the accuracy of predictions. The growth of the secondary battery sector can be attributed to global environmental changes such as worldwide energy transition policies, carbon neutrality goals, and an accelerated shift toward renewable energy technologies. This study concludes that using posts from stock discussion forums directly shaped by individual investors enhances the sophistication of predictions. Accordingly, theoretical and practical implications are put forth.
AB - The advent of social media has transformed the stock investment landscape considerably. Individual investors are increasingly shifting from traditional financial institutions to platforms such as Facebook, X, and stock discussion forums to obtain information and exchange opinions, significantly affecting the stock market. Post-COVID-19, the influx of individual investors has accentuated their importance in the South Korean stock market, necessitating the development of prediction models utilizing data from stock discussion forums, a facet of social media. Research on predicting stock price fluctuations using social media data is lacking, and most studies have focused on news articles, with insufficient utilization of stock discussion forum posts. Considering the significant proportion of individual investors in the Korean stock market, it is imperative to conduct research utilizing posts from stock discussion forums, where individual investors' opinions are directly formed, for accurate stock price predictions. This study specifically focuses on the secondary battery sector, which garnered significant attention during the pandemic, to examine the sentiments of individual investors. Employing data from stock price prediction models, this study compares the accuracy of predictions. The growth of the secondary battery sector can be attributed to global environmental changes such as worldwide energy transition policies, carbon neutrality goals, and an accelerated shift toward renewable energy technologies. This study concludes that using posts from stock discussion forums directly shaped by individual investors enhances the sophistication of predictions. Accordingly, theoretical and practical implications are put forth.
KW - Individual investor
KW - secondary battery sector
KW - sentiment analysis
KW - stock discussion forum
KW - stock price prediction
KW - topic modeling
UR - http://www.scopus.com/inward/record.url?scp=85207722283&partnerID=8YFLogxK
U2 - 10.1109/ACCESS.2024.3481442
DO - 10.1109/ACCESS.2024.3481442
M3 - Article
AN - SCOPUS:85207722283
SN - 2169-3536
VL - 12
SP - 153537
EP - 153549
JO - IEEE Access
JF - IEEE Access
ER -