Machine Learning Based SPAM Message Classification System using Blockchain Technology

Abstract

Short Message Service (SMS) is one of the most common communication media for mobile devices. Although there are many alternatives to mobile text communication such as Whatsapp or LINE based online, SMS is still very commonly used because it only requires mobile number information and mobile signals available. Because of the ease of disseminating information with this SMS media, many people use SMS to send unimportant or even harmful messages to mobile device users. This unimportant SMS is referred to as SPAM. The number of SPAM messages that enter a person’s mobile phone filled the inbox with non-essential information, covered important information, wasted user time, and hindered communication. Therefore, a system that can automatically classify SPAM is required. Data obtained from this classification system may contain sensitive user information such as mobile numbers. Therefore, a SPAM SMS classification storage system with a blockchain platform is created. In this work, we use Hyperledger Composer to create our blockchain. Blockchain technology is used because it is an immutable and secure data storage system that can maintain the confidentiality of sensitive information that may be contained in SPAM messages.