Since the invention of the internet and the revolution that followed, the scope of fraud and fraud detection has evolved completely. Fraud attacks are constantly increasing in sophistication and costing worldwide financial institution billions of dollars annually.
With technology-driven fraud on the rise, organizations need to conduct a structured and comprehensive fraud risk analysis to unearth limitations and identify possible risks they may be susceptible to.
This fraud risk analysis is essential to aid organizations with the development of an appropriate risk management framework that best utilizes the efforts of both the Anti-fraud and IT security teams.
Taking a holistic approach by integrating fraud prevention processes and cybersecurity technology, possibly creating a fusion center, is the most feasible way to address the increasingly technical nature of fraud attacks.
Harnessing technology plays a decisive role in both attack and defence. Staying ahead in this arms race between organizations and cybercriminals means taking advantage of emerging technology to not only mitigate attacks but prevent them before they occur.
Linking multiple data sources and applying an analytics-based approach like using descriptive and predictive analytics, machine learning and social network analysis to analyze the large volume of data extracted from these sources is one of the newer methods for predictive fraud detection.
This approach is simply applying data mining techniques to fraud detection by using multiple data sources to identify potentially fraudulent patterns.
Social Network Analysis (SNA) specifically, is one of the prevailing data mining methods in fraud analytics today. Social Network Analysis goes beyond just social media, it is a connected network of entities. Entities could be ranging from companies to fraudsters, online transactions, social media data and call behaviour data etc which are usually stored in unstructured form in environments like social media, payment gateways and bank servers.
Graph database technology can be specially modified to explore these large networks of connections and relationships and detect suspicious patterns of behaviours which it then stores in native network graph format to help organizations discover other hidden structures, locate clusters and patterns and identify links in transaction chains.
Financial crime against banks and other financial institutions is accelerating rapidly. Banks need to improve their defences by upgrading and optimizing their transaction system and fraud management tools, also incorporate new predictive technology in their security framework to reduce these losses.
Any Fraud management tool considered for integration should be able to:
- Enhance information credibility
- Detect fraud faster with real-time integration
- Improve behaviour monitoring of individuals
- Uncover hidden relationships and subtle patterns of behaviour
The role of predictive technology in combatting bank frauds has changed the very way we react fraud risks in most financial institutions and organizations but there’s still a long way to go to ensure the safety of financial data and assets. Otis Yeboah is the Snr. Financial Crime Analyst Financial Crime Management of Fidelity Bank Ghana Ltd. and he will be sharing his knowledge on the growing fraud attack landscape at the African Cybersecurity and Fraud Prevention Forum held from May 6-7, 2020 in Lagos, Nigeria.