Skip to content

Latest commit

 

History

History
12 lines (7 loc) · 381 Bytes

README.mdown

File metadata and controls

12 lines (7 loc) · 381 Bytes

Bayesian Spam Filter

We implement a Bayesian spam filter in Python.

We first implemented the vanilla Bayesian filter, training on the data provided, and then extended it in various directions.

Extensions

  • Stemming using Porter's algorithm
  • Cosine normalisation of feature weights
  • Feature selection using WEKA.