First, we need to get a large collection of users and their locations with a broad range of interests. We can then write a script to determine the number of twitter users we have collected per city (regardless of the interest they're counted in). We can then compare for each interest the number of users in a city that we retrieved for that interest against the percentage of twitter users in a city.
First, we need to get a large collection of users and their locations with a broad range of interests. We can then write a script to determine the number of twitter users we have collected per city (regardless of the interest they're counted in). We can then compare for each interest the number of users in a city that we retrieved for that interest against the percentage of twitter users in a city.