(For Staff & Students only)
NUS NoSpamMail employs heuristic technology that evaluates and identifies e-mails using multiple levels of Spam characteristics. Its heuristic rules work like artificial intelligence to produce a Spam confidence rating. It begins with a baseline detection rate and uses tune-able preferences to achieve accuracy to recognize definition of spam.
What Is Spam?In general terms, Spam is considered to be any unsolicited e-mail that often violates the normal acceptance criteria for desired communications. Spam typically includes commercial or money-making offers, but can also include messages with socially objectionable content related to pornography, racially sensitive discussions, or other offensive topics.
A key problem in Spam detection is that not all Spam are obvious; Some messages that appear to be Spam are in fact legitimate e-mail, and vice-versa. Any methodology that categorizes Spam must consider organizational or user interpretation when evaluating the grey area of suspicious content.
A newsletter describing a new way to make money might appear to be Spam to most people; however, to a person looking for a new career, this can be regarded as a promising opportunity. Spam detection technologies must consider both the success rate and the accuracy to be a sound Spam detection system.