A Robust Approach for Effective Spam Detection Using Supervised Learning Techniques

Abstract

A collection of millions of devices with sensors and actuators that are linked via wired or wireless channels for data transmission. Over the last decade, it has grown rapidly, with more than 25 billion devices expected to be connected by 2020. The amount of data released by these devices will multiply many times over in the coming years. In addition to increased volume, the device generates a large amount of data in a variety of modalities with varying data quality defined by its speed in terms of time and position dependency. In such an environment, machine learning algorithms can play an important role in ensuring biotechnology based security and authorization, as well as anomalous detection to improve usability and security. On the other hand, attackers frequently use learning algorithms to exploit system vulnerabilities. As a result of these considerations, we propose that the security of devices be improved by employing machine learning to detect spam. Spam Detection Using Machine Learning Framework is proposed to attain this goal. Four machine learning models are assessed using multiple metrics and a vast collection of input feature sets in this framework. Each model calculates a spam score based on the input attributes that have been adjusted. This score represents the device's trustworthiness based on a variety of factors. In comparison to other current systems, the findings collected demonstrate the effectiveness of the proposed method.