I’m looking for the proper way to get a representative sample of tweets written in a determined language. I view two options here:
- Use the filter endpoint and track the most popular stopwords in this language. I guess I would get up to 1% of tweets written in the language.
- Use the sample endpoint with the language paramater. I can’t guess the amount of tweets I would get here.
What is the recommended way to go to get a bigger and representative sample?