I am currently developing a Twitter bot that identifies suicidal and depressed tweets and replies with emergency crisis hotline information should they need it. I am doing some sentiment analysis on the tweets after I get them, but I am utilizing the stream API’s track attribute to do a preliminary filter and get only the Tweets that may be applicable. I am doing this by combining a list of phrases with each of them containing a trigger word, like “kill” or “suicide”, a personal pronoun, and a desire or action word like “commit”, “want”, and “wish”. There are more but I’m trying to keep it short. The problem is that you multiply the number of trigger words I’m tracking by the number of personal pronouns, and then multiply that by the number of desire words, and the size of the list grows really quick. I added “depressed” to the trigger word list and it push me to 450 tracked phrases; over the 400 limit, so now if I think of a word to add to the lists, I get a 413 response error.
However, I saw on the API response code page that I could contact the API team for increased access. This is a simple personal project of mine, and will run on a Raspberry Pi in my room, so there are no other users and it isn’t anything like an iPhone application, so I was just wondering how easy of a process it may be to get an increased limit.