Skip to content

Tweet classifier in Haskell distinguishes aggressive and non-aggressive tweets

License

Notifications You must be signed in to change notification settings

Pold87/Haggressive

Repository files navigation

Tweet classifier in Haskell

Idea

Tweets on Twitter may contain aggressive content. Haggressive is a supervised k-Nearest Neighbor machine learning classifier that distinguishes aggressive and non-aggressive tweets. It uses a bag-of-unigram approach for extracting features. The empirical evaluation of the approach has yielded an average accuracy of 68.57 % of k = 10. See report_haggressive.pdf for the accompanying report.

Usage

Run it with cabal and specify the directory for the dataset. The actual main file is in src-lib/Hag.hs.

About

Tweet classifier in Haskell distinguishes aggressive and non-aggressive tweets

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published