TY - JOUR
T1 - Exploring the sentimental features of rumor messages and investors’ intentions to invest
AU - Alzahrani, Ahmed Ibrahim
AU - Sarsam, Samer
AU - Al-Samarraie, Hosam
AU - Alblehai, Fahad
PY - 2023/9
Y1 - 2023/9
N2 - Investors' decisions to invest in stock markets can be significantly influenced by online rumors generated by certain companies or influencers. The current understanding of how certain sentimental features can help increase the prediction capabilities of online rumors is still in its fancy stage. This study explored the types of topics and emotions found in rumor messages and how they are associated with investors' decisions to invest in stocks. We also investigated the potential of using these emotions in predicting investors' intention to invest in stock markets. The sentimental features consisted of users' emotions (anger, fear, sadness, joy, and trust) and polarity (positive, negative, and neutral). A topic modeling approach was applied to identify logical associations between different sentimental features of rumors on Twitter. The results showed that rumors tweets associated with investors' intention to invest were linked to the joy and trust sentiments, while the anger and fear sentiments were linked to no intention to invest. The results showed that these emotions can be used in predicting the impact of online rumors on investors’ investment decisions. The prediction model can be useful for stock market prediction by enabling managers and researchers to analyze and assess the magnitudinal impact of rumors on certain investment decisions. The outcomes can also help decision and policy makers to take the required actions to prevent possible financial instability due to COVID-19 or other future events.
AB - Investors' decisions to invest in stock markets can be significantly influenced by online rumors generated by certain companies or influencers. The current understanding of how certain sentimental features can help increase the prediction capabilities of online rumors is still in its fancy stage. This study explored the types of topics and emotions found in rumor messages and how they are associated with investors' decisions to invest in stocks. We also investigated the potential of using these emotions in predicting investors' intention to invest in stock markets. The sentimental features consisted of users' emotions (anger, fear, sadness, joy, and trust) and polarity (positive, negative, and neutral). A topic modeling approach was applied to identify logical associations between different sentimental features of rumors on Twitter. The results showed that rumors tweets associated with investors' intention to invest were linked to the joy and trust sentiments, while the anger and fear sentiments were linked to no intention to invest. The results showed that these emotions can be used in predicting the impact of online rumors on investors’ investment decisions. The prediction model can be useful for stock market prediction by enabling managers and researchers to analyze and assess the magnitudinal impact of rumors on certain investment decisions. The outcomes can also help decision and policy makers to take the required actions to prevent possible financial instability due to COVID-19 or other future events.
KW - Social media
KW - Investment
KW - Decision making
KW - Topic modelling
KW - Sentiment analysis
KW - Emotions
UR - http://www.scopus.com/inward/record.url?scp=85160313627&partnerID=8YFLogxK
U2 - 10.1016/j.iref.2023.05.006
DO - 10.1016/j.iref.2023.05.006
M3 - Article
SN - 1873-8036
VL - 87
SP - 433
EP - 444
JO - International Review of Economics and Finance
JF - International Review of Economics and Finance
ER -