Preventing Next-Generation Fraud with Graph Database Technology

Annual losses from cybercrime worldwide are estimated to reach $10.5 trillion by 2025. These losses impact everyone from financial institutions, to businesses and individuals — victims of fraud.

Part of the problem arises from the fact that many accounts rely on little more than a username and password to protect access, simplifying account takeover and identify theft. Another facet is the surge in overall credit card transactions, accelerated by the pandemic. This is exacerbated by increased reliance on mobile devices to perform financial transactions. When combined with social engineering, scams become even more challenging to prevent. 

Preventing Fraud Real-time

The first challenge in overcoming fraud is to identify that it’s taking place. The second is to stop it before it can cause significant damage. Better yet would be to prevent fraud from happening in the first place.

Identifying and stopping fraud requires a layered intelligent approach. This means that systems need to look beyond transactions to relationships including verifying identification based on a holistic view of actions and behaviors.

Uncharacteristic purchases can be a relatively straightforward way to identify fraud because the credit card company completing the transaction has immediate access to transaction history. But what if the scammer makes these purchases using different credit cards? Identifying more complex fraud requires a whole new way of looking at data.

Graph Databases Map the Relationships

Traditional relational databases are limited at capturing the relationships between data points. The confines of tabular data models and rigid schemas make it difficult to add new or different kinds of connections.

Graph databases enable new approaches to identifying fraud and complex scams with a remarkable level of accuracy. They achieve this through advanced contextual link analysis. The more data points that can be linked and interconnected, the more holistic the view of patterns, people and their behavior. 

Once fraud is identified, it can be stopped in real time. And, as the graph database system learns the signs of fraud, similar fraudulent operations can be learned, detected and stopped faster. Say five transactions using different cards have been made from the same IP address. This goes beyond what might be considered reasonable behavior for an authentic customer. The action can be flagged as potentially fraudulent and escalated to a higher level of authentication such as sending a message to every cardholder verifying the purchase, and/or immediately blocking all the transactions. 

The graph database system can also integrate this patten and automate response to it through artificial intelligence. Fraud detection and user verification can occur real-time. Note that this applies to any identifying information, not just an IP address. If a specific email or login credential is identified in a potentially fraudulent transaction, all future interactions with this email or credential can be subject to additional scrutiny.

Real-time Fraud Detection

New Day is a financial services company specializing in consumer credit products in the UK. Given that New Day issued 19% of new credit cards in the UK In 2020, they are a prime target for application fraud. Their legacy databases could not analyze the volumes of data fast enough to block fraudulent activities without negatively impacting their responsiveness. 

Deep link analytics combined with real-time analysis and machine learning provide a robust platform for detecting and preventing fraud. When New Day’s investigators identify new patterns of fraud, they can now share this information with other investigators. They can also integrate it into their systems to identify similar cases going forward. The immediate result was a 10-15% reduction in fraud.

Another financial institute wanted to improve its networking and link analysis capabilities to detect money laundering. With a graph database solution, the firm can display the connections from an ecosystem surrounding the point of interest and identify which connections would most likely lead to productive investigations. With this system in place, analysts can now react to potential money laundering work items in real-time.

Beyond Fraud to Improving Recommendations

Graph database technology can provide benefits beyond fraud detection. Other important applications for graph databases include:

• Real-time recommendation engines: Accurately offering options relevant to a specific user increases the chance of engagement.
• Master data management: Master data, such as a business’ customers, products and suppliers, is dynamic in nature and includes many relational aspects that are difficult to identify and manage with a traditional database.
• Network and IT operations: The physical and human interdependencies of networks are extremely complex, making it difficult to scale operations and troubleshoot issues.
• Identity and access management: To verify an accurate identity and its access permissions, systems must traverse a highly interconnected dataset that is constantly growing in size and complexity, impacting responsiveness.

With the ability to analyze the interconnected relationships of data, organizations can detect and prevent even complex fraud and scams in real-time. Dell Technologies is at the forefront of analytics to help our customers leverage innovative technology such as graph databases. 

The world is only getting more complex. With the right technology, your organization can keep ahead of fraud. Find out more and take a test drive in one of Dell Technologies’ worldwide Customer Solution Centers.

***

Intel® Technologies Move Analytics Forward

Data analytics is the key to unlocking the most value you can extract from data across your organization. To create a productive, cost-effective analytics strategy that gets results, you need high performance hardware that’s optimized to work with the software you use.

Modern data analytics spans a range of technologies, from dedicated analytics platforms and databases to deep learning and artificial intelligence (AI). Just starting out with analytics? Ready to evolve your analytics strategy or improve your data quality? There’s always room to grow, and Intel is ready to help. With a deep ecosystem of analytics technologies and partners, Intel accelerates the efforts of data scientists, analysts, and developers in every industry. Find out more about Intel advanced analytics.