Classifying cyberattacks on financial organizations based on publicly available deep web dataset

Abstract

The financial sector is one of the main targets of malicious persons stealing money and data from individuals and organizations. The lack of knowledge of end users, expertise of attackers, and availability of tools and publicly accessible data make it serious about keeping all online transactions safe. Hence, it causes massive damage in terms of data and fund loss. Surprisingly, the attackers make this database public to show their ability, threaten a specific community, or find potential buyers on the dark web. However, studying this dataset reveals the current situation and upcoming threats in a broader picture. Moreover, it may help to take precautions for subsequent potential attacks. The study solely focuses on data sets for both surfaces web - 'Paysim' and 'Banksim,' and dark web - SOCRadar dataset to analyze cyberattacks mainly targeting transactions. It also classifies cyberattacks on financial organizations. The research team finds there are mainly 71 cyberattacks that are generated around the world. Most cybercriminals mainly target the United States, Russia, China, the United Kingdom, Spain, and Germany. However, the US ranks at the top stated-backed hackers based on the dark web dataset. The study also shows an overview of cyberattacks based on transaction type, gender, and fraud distributions. The team uses the Python-based Beautiful Soup package to scrap web data. The team highly analyzes the datasets and visualizes using Seaborn. The research team also proposes a comprehensive four-step framework, FinFrame, for securing financial organizations. Finally, the team suggests ensuring a minimum ground level of knowledge in terms of cybersecurity for the end users to keep safe from being hacked.