期刊名称：International Journal of Advanced Research In Computer Science and Software Engineering
印刷版ISSN：2277-6451
电子版ISSN：2277-128X
出版年度：2013
卷号：3
期号：3
出版社：S.S. Mishra
摘要：In recent years, Na.ve Bayes classifier has gained much popularity in spam classification due to its simplicity and superior performance. We studied the performance of na.ve bayes classifier and found that it largely depends on the smoothing method, which aims to adjust the probability of an unseen event from the seen event, that arises due to data sparseness. Therefore in this paper, we aim at enhancing the performance of na.ve bayes classifier in classifying spam mails by proposing a modification to Absolute Discount smoothing method against the laplace method of traditional na.ve bayes classifier. In addition, we have introduced a cost metric to co mpare our approach with the traditional scheme. Our experimental results have shown that our method not only achieves grea ter accuracy as compared to Laplace but also reduces false positives, which is more serious problem in spam classification
