期刊名称：International Journal of Innovative Research in Computer and Communication Engineering
印刷版ISSN：2320-9798
电子版ISSN：2320-9801
出版年度：2014
卷号：2
期号：12
出版社：S&S Publications
摘要：Spam is email sent in bulk wherever there's no direct agreement in situ between the recipient and alsothe sender to receive email solicitation. to forestall the delivery of this spam, an automatic tool named a spam filter isemployed. during this paper, (OBP) “Optical Back Propagation” technique is employed as an automatic tool to spotwhether or not a message is spam or not supported the content of the message. Spam-based dataset-a dataset from UCI“University of California, Irvine” machine learning repository, is employed as coaching and testing dataset to coach thenetwork so tested it. The samples of this dataset ought to be first of all preprocessed (normalization or feature choicebefore normalization) to be appropriate to the network. The results OBP spam filtering is affordable in term ofaccuracy, precision, recall, false Positive, false negative, and speed of net.
关键词：Neural Network; Optical Back Propagation; Spam; Spam Filter
