This is an ongoing exercise using the Twitter API to gather tweets about Cafeyn brands - Cafeyn, Blendle and MiLibris. A cron job is running on a Raspberry pi every day to collect new tweets about the company and store them in a csv.
Visualizing results
The second part is done in R and ggplot. The teets are analyzed and outputed into static charts that can be uploaded on a website. Ideally, we could have a setup where the graphs are hosted on the Raspberry via Shiny, and a website could access them in a dynamic way.
Next steps
- Have a minimal working example of a setup where the tweets are gathered, processed and outputed on a page from my website
- Make the graphs dynamic, probably hosting the stored file on S3 or GCS to learn that setup too
- Extend Twitter API calls to a wider spectrum than just Twitter search results
- Expand analysis onto NLP when there is more content