Twitter sentiment analysis: An Arabic text mining approach based on COVID-19.

Area of Science:

• Natural Language Processing
• Public Health Communication
• Social Media Analytics

Background:

• Social media platforms have become primary channels for information dissemination and public discourse.
• The COVID-19 pandemic exacerbated the spread of misinformation and public anxiety, necessitating methods to gauge public sentiment.
• Analyzing public emotions expressed on social media is crucial for understanding societal needs and informing policy.

Purpose of the Study:

• To develop and evaluate a sentiment analysis model for detecting genuine news related to the COVID-19 pandemic in Arabic text.
• To analyze public sentiment towards the COVID-19 pandemic in Gulf countries using Twitter data.
• To provide a tool for authorities to understand public emotions and inform public health strategies.

Main Methods:

• Utilized a sentiment analysis model incorporating Machine Learning techniques.
• Employed the Synthetic Minority Over-sampling Technique (SMOTE) to address imbalanced datasets.
• Focused on Arabic text data from Twitter specifically from Gulf countries during the COVID-19 pandemic.

Main Results:

• The sentiment analysis model successfully processed Arabic text data from Twitter.
• Analysis indicated a predominantly negative public sentiment among people in Gulf countries during the COVID-19 pandemic.
• The findings underscore the utility of sentiment-based data mining for understanding public reactions during health crises.

Conclusions:

• Sentiment analysis of social media data provides valuable insights into public emotions during health crises like the COVID-19 pandemic.
• Government authorities can leverage this approach to directly understand public sentiment and implement targeted interventions.
• The developed model offers a method for monitoring and mitigating the impact of misinformation and public anxiety.