Some notes on what we are trying to accomplish:
We are aiming to find a correlation between mental health decline and COVID sentiment (per country) We will accomplish this by:
- adding alcohol consumption data
-normalize it based on country vs amount of tweets (take a percentage)
-alcohol vs negative tweets - correlation?? - other things that are attributed to lower mental health:
-suicides
-substance abuse (opiod)
-crime
-depression
Possible prototype implementations:
- interactive heatmap of [variable] vs COVID sentiment
- word cloud (neutral vs negative vs positive)
- app interface