I collected 64 e-mails from DBWorld newsletter and I used them to train different algorithms in order to classify between 'announces of conferences' and 'everything else'. I used a binary bag-of-words representation with a stopword removal pre-processing task before.

Attribute Information:

Each attribute corresponds to a precise word or stem in the entire data set vocabulary (I used bag-of-words representation).

Relevant Papers:

Michele Filannino, 'DBWorld e-mail classification using a very small corpus', Project of Machine Learning course, University of Manchester, 2011. [Web link]

Citation Request:

Thanks to ACM-SIGMOD for its useful service! :)

Michele Filannino, PhD
University of Manchester
Centre for Doctoral Training
Email: filannim_AT_cs.man.ac.uk