Home

About

Stream tweets which mention COVID19 related terms and have their user location non-empty, and analyse the sentiment per area.

The team

The team consists of:

• Ioannis Pavlopoulos
• Giannis Panagiotakis
• Ilias Tsoumas
• Ion Androutsopoulos

Completed steps

• Deployed to server and saving tweets.
• Initiated annotation on emotions (based on this article) on Greek tweets.
• Trained emotions detection classifier.

Next steps

• Set up a site to show live a map with sentiments/emotions over time.
• Deploy emotions detection system.
• Show emotions on map.

Deliverables

• D1: COVID19-specific sentiment/emotions classifier for english and greek tweets.
• D2: Site to monitor the sentiment/emotions per area across time.

Milestones

• M1: Start saving all the COVID19-related tweets [expected by April 15].
• M2: Complete an annotation task of 2K tweets about COVID19 (1K per language).
• M3: Assess several sentiment/emotions classifiers.
• M4: Build new systems that will outperform the baselines in both languages.