A Data Analytics Approach to the Cybercrime Underground Economy

The number of cyber threats is increasing very fast all over the world. This creates serious problems for individuals, companies, and governments. Even though cybersecurity has become more important in recent years, there is still a lack of proper research and clear methods to help people understand and deal with cybercrime effectively. One important concept in this area is Crime-as-a-Service (CaaS). This means that cybercriminals provide hacking tools, services, and systems to others in exchange for money. Because of this, even people with very little technical knowledge can carry out cyberattacks. As a result, cybercrime has become more organized, scalable, and easy to access. Today, there is a large underground market where services like malware creation, phishing kits, ransomware attacks, and stolen data are bought and sold. CaaS has also made cybercrime more professional. Different people in this system have different roles. For example, some create malicious software, others spread it, and some sell stolen data. This division of work makes cybercrime more efficient and powerful. Also, cybercriminals use technologies like encryption and anonymous networks, which makes it very difficult for law enforcement agencies to track them. This project focuses on studying the cybercrime underground economy using data analytics and a design science research method. It proposes a structured framework to collect, process, and analyze data from hacking forums and underground platforms. The system uses data preprocessing techniques such as cleaning, normalization, and feature selection to improve data quality. This helps in converting raw data into useful information and understanding cybercrime patterns better. In addition, the project clearly defines important terms like CaaS and crimeware to improve understanding for both researchers and professionals. Based on these concepts, a classification model is developed using the Naive Bayes algorithm. This algorithm is simple, fast, and works well with large datasets. It analyzes patterns in the data and classifies different types of cybercrime activities, making it very useful for studying text-based and behavioral data.