SPAMMER DETECTION USING ENHANCED GAUSSIAN MIXTURE MODEL IN MOBILE CLOUD ENVIRONMENT
There are several non-automated spam detection structures available that do not rely on machine learning policies. Still, these systems suffer significant bottlenecks (blocks) in throwing (attempting) present (modern) spam attack models and dynamism. Spam content is growing with an unstable development of user-generated content (UGC) on the web. Spammers frequently include popular keywords or copy and paste original commodities from the internet with spam links fixed, endeavoring to incapacitate content-based discovery. With the majority of the mobile network society, spammers have made into clusters for profit maximization, which has produced excitement and considerable damages to manufacturing generation. It is challenging to detect spammers from regular users owing to the features of multi-dimensional data. To discuss this difficulty, this paper offers a Spammer Identification Mechanism using Enhanced Gaussian Mixture Model (EGMM) that employs machine learning for modern mobile interfaces. It performs intelligent identification of spammers without relying on manageable and unstable connections. The proposed scheme merges the performance of information, where every user node is divided into one class in the development method of the representation. We verify EGMM by comparing it with the K-means clustering (KMC) and hybrid FCM clustering (HFCM) mechanism utilizing a mobile network dataset from a cloud server. The model has been more distinguished toward similar investigations, and the result shows that the proposed system results in an increase of 10-15% in the accurate detection rate of spam and harm.