Data365 API for Monitoring Public Opinion on COVID-19 Epidemic Response Measures

1. Data collection (Data365 Social Media API)
   
   • The project began with the automated collection of public posts from several social media networks using the Data365 API, which allowed our client to efficiently gather relevant data.
   • Utilizing over 360 unique search queries related to coronavirus and government topics, Data365 facilitated the extraction of 6,726 posts for the period from April to May 2020, focusing on users located in a specific area.
2. Preprocessing of data
   
   • The collected data was processed by formatting text, converting it to lowercase, removing unnecessary characters and stop words, and applying stemming techniques.
3. Semantic analysis of text
   
   • Using tools based on Word2Vec, semantic analysis of the text corpus was conducted.
   • The Word2Vec model helps to understand the relationships between words that frequently appear together, allowing the identification of key narratives, such as discussions about fines and violations in connection with quarantine, and government efforts against the pandemic.
4. Sentiment analysis of users
   
   • Additionally, the client developed sentiment dictionaries in order to analyze the emotional tone of posts. The results showed that negative emotions relating to the way the government handled things, especially on words like ‘government’, ‘quarantine’ and ‘president’ dominated.
5. Reporting and visualization of results
   
   • The client utilized detailed reports, enriched with visualizations showing user ‘sentiment’ distribution, to present their findings. The insights were clear about public opinion, and knowledge was critical to evaluating the impact of government communication strategies.

Key Benefits

• Efficient data collection: The Social Media API of Data365 made it very easy and precise to collect the required posts to deliver the client with the comprehensive dataset.
• Informed decision-making: Using Data365 to provide the real time data allowed the client to make informed decisions as we kept the public sentiment up to date.
• Enhanced communication strategies: The analysis of the data provided the insights used by the client to improve their communications to better respond to public concerns.