In this paper, we study the email classification problem. We apply the notion of shingling to capture the concept of phrases. For each email, we form a sketch which is compact in size and the sketch of two emails allows for computation of their resemblance. We then apply a k-nearest neighbour algorithm to classify the emails. Experimental evaluation shows that a high degree of accuracy can be obtained.
CITATION STYLE
Poon, C. K., & Chang, M. (2003). An email classifier based on resemblance. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 2871, pp. 344–348). Springer Verlag. https://doi.org/10.1007/978-3-540-39592-8_48
Mendeley helps you to discover research relevant for your work.