Rough Set Approach to Spam Filter Learning
Mawuena Glymin, Wojciech Ziarko · Lecture notes in computer science · 2007
This article presents an elementary overview of techniques employed for spam detection via probabilistic decision table-based predictive data modelling. The focus here is to present a solution that combines simple algorithms together with some heuristics to construct generalized rough approximations of spam and legitimate e-mails using the variable precision rough set (VPRSM) approach. Experiments were conducted to explore the application of VPRSM for designing an intelligent agent for spam filtering.