Ekonomi & Bisnis editorial
Open AccessOA2023
Analyzing influence of COVID-19 on crypto & financial markets and sentiment analysis using deep ensemble model
This study analyzes COVID-19's impact on public sentiment toward financial and crypto markets using tweets, proposing a CNN-LSTM model that achieves scores of 0.89 and 0.92 for crypto and financial markets, respectively. It also performs damage analysis and topic extraction to understand investment sentiments during the pandemic.
P. Washington; Pradeep Gali; F. Rustam; Imran Ashrafยท PLoS ONEยท 2023ยท DOI 10.1371/journal.pone.0286541
The core problem
The COVID-19 pandemic severely affected the global economy, increasing inflation in both developed and developing countries. Financial and crypto markets were significantly impacted, yet some crypto markets flourished and reached their peak during the pandemic. This study investigates the impact of COVID-19 on public opinion and sentiments regarding financial and crypto markets by conducting sentiment analysis on tweets posted during peak COVID-19 days. It explores people's sentiments about investment in these markets and performs damage analysis in terms of market value, identifying the worst times for these markets. The proposed approach uses a hybrid CNN-LSTM model for sentiment classification and topic extraction to understand sentiments related to each topic.
Innovation
Data was extracted from Twitter using the SNSscraper library, focusing on tweets related to financial and crypto markets during the COVID-19 peak days. The study proposes a hybrid model called CNN-LSTM (convolutional neural network-long short-term memory) for sentiment classification. The model combines CNN for feature extraction and LSTM for sequence learning. The dataset was preprocessed and split into training and testing sets. The model was trained and evaluated using score as the primary metric. Additionally, topic extraction was performed on the tweets to identify prevalent topics and their associated sentiments. Damage analysis was conducted by examining market value changes over time to determine the worst periods for financial and crypto markets.
Introduction
The COVID-19 pandemic severely affected the global economy, increasing inflation in both developed and developing countries. Financial and crypto markets were significantly impacted, yet some crypto markets flourished and reached their peak during the pandemic. This study investigates the impact of COVID-19 on public opinion and sentiments regarding financial and crypto markets by conducting sentiment analysis on tweets posted during peak COVID-19 days. It explores people's sentiments about investment in these markets and performs damage analysis in terms of market value, identifying the worst times for these markets. The proposed approach uses a hybrid CNN-LSTM model for sentiment classification and topic extraction to understand sentiments related to each topic.
Why it matters
The results demonstrate the effectiveness of the CNN-LSTM hybrid model in sentiment classification for financial and crypto markets during the COVID-19 pandemic. The high scores suggest that the model can accurately capture public sentiment from tweets. The topic extraction provides insights into the specific concerns and interests of investors during the crisis. The damage analysis offers a quantitative view of the pandemic's impact on market values, identifying periods of severe decline. The study highlights the dual nature of the pandemic's effect: while some crypto markets flourished, others experienced significant damage. The findings can inform investors and policymakers about market sentiment dynamics during global crises. Future work could extend the model to other social media platforms and real-time analysis.
Who should read this
Pembaca Research Digest dan praktisi rumpun ini.
Opening member contentโฆ