Chapter

Information Security Applications

Volume 3325 of the series Lecture Notes in Computer Science pp 135-143

Some Fitting of Naive Bayesian Spam Filtering for Japanese Environment

  • Manabu IwanagaAffiliated withGraduate School of Information Science and Electrical Engineering, Kyushu University
  • , Toshihiro TabataAffiliated withFaculty of Information Science and Electrical Engineering, Kyushu University
  • , Kouichi SakuraiAffiliated withFaculty of Information Science and Electrical Engineering, Kyushu University

* Final gross prices may vary according to local VAT.

Get Access

Abstract

Bayesian filtering is one of the most famous anti-spam measures. However, there is no standard implementation for treatment of Japanese emails by Bayesian filtering. In this paper, we compare several conceivable ways to treat Japanese emails about tokenizing and corpus separation. In addition, we give experimental results and some knowledge obtained by the experiments.