Analysis and detection of SPIM using message statistics

SPam over Instant Messaging (SPIM), is becoming one of the major challenges for the IM Networks. These messages affect the overall performance in various ways, ranging from waste of resources e.g. Memory, unwanted traffic on the Network, to major disasters e.g. spread of worms, DoS attacks etc. Different techniques have been proposed for the detection of SPIM; based on content-based searches, feature extraction etc. In this paper a novel technique for detection of SPIM has been discussed. An in-depth analysis of SPIM and proof of the effectiveness of the technique based on the results gathered from a real data set has also been provided.