I’ve been playing around with Twitter’s API and decided to take a shot at analyzing AniTwitter. If you don’t know what that is, look it up on Urban Dictionary or read below.
First, let’s take a look at which hashtags were the most common among the users on my AniTwitter list.
I also considered adjectives, to see whether people were feeling positive or negative.
Last but not least, these tweets got the most retweets.
Alexis (@Alexis_Bich) May 23, 2015
When your priorities are right http://t.co/mcfE9thEw9—
Ashlynn (@AshlynnArias) May 26, 2015
いしがな (@nobita_ishigana) May 21, 2015
Just looking at pictures for Mad Max and found this, oh mygod http://t.co/HwArgXMsKE—
Callie (@cfzelda) May 25, 2015
I have ramblings about that Shirou pic, but I’ll save them for another time.
So, what is AniTwitter? AniTwitter is a loosely-connected community of Twitter users that like to talk about anime-related stuff, among other things. Since it would be kind of presumptuous to just come up with a list of users that I consider to be part of AniTwitter, I looked through several publically available curated lists from other users and merged them into a mega list of 3700 unique Twitter accounts. It’s not a definitive list, but more than enough usual suspects are on it. The list includes mostly English speakers. Most of them are fans, some are company accounts. I have yet to publish the list on Twitter, but I’ll do so when I have time. Expect another post like this one next week.
During last week, people have posted over 360 thousand tweets, retweets included. I used tweepy, a Pythonian RESTful Twitter API library, to harvest them from Twitter. I had to override some classes, to get more out of the API than was defined, but overall I’m quite happy with it. I used Natural Language ToolKit for analyzing word forms, and Word Cloud for generating word clouds.
As always, happy to take suggestions and stats requests. If you want the code, hit me up on Twitter.