Sentiment analysis throughout the pandemic period
Sentiment analysis of COVID-19 related tweets in Portuguese throughout the pandemic.
- Github Link
- july 2020 - october2020
- Delivered
Objectives
Creation of a visual timeline that reflects the emotional state of the Portuguese population throughout the pandemic period and integrates information on the measures that may have contributed the most to that state.
Results
1 – Selection of relevant data sources for sentiment analysis of the Portuguese population (such as data from our news repository Observador/Público newspaper, or data extracted from Twitter). – List implementation and survey of government measures over time;
2 – Development/use of pre-existing algorithms (e.g. NLTK/spaCy in Portuguese) for classification in terms of positive/negative sentiment (or a range in between), based on text from the chosen sources;
3 – Presentation of the results through relevant and explicit visualizations;
More detail in the blog post written by the volunteers: https://www.dssg.pt/a-pandemia-nos-nossos-sentimentos/