TigerGraph has released its new technology which combines graph pattern matching with real-time deep link analytics to boost fraud detection and cyber protection.
By releasing this new software, the company is hoping to improve the ability for enterprises to generate deep insights from their data, and improve anti-fraud, AML, security analytics, recommendation engines and AI capabilities.
Pattern matching technology has been typically held back by its inability to scale computational requirements on large datasets, and difficulty linking analytics, the company said. An example given, was determining the ultimate beneficiary ownership in banking and financial services would entail finding its parent company, finding its stakeholders for each division and adding up ownership proportions, with each step needing parallel computation to traverse data, which is tough to do.
When trying to detect fraud or money laundering incidents, it is also tough to originate where it came from and find the source, due to fraudsters often hiding operations under multiple synthetic accounts, TigerGraph said.
To combat this, the TigerGraph solution is able to monitor these hidden operations, or hops, to identify the origins.
TigerGraph CEO and founder Dr. Yu Xu said, “Efficient graph analytics is more than just a great massively parallel processing engine; it’s understanding what users want to know and focusing on that and pruning away the rest. The communicative and intuitive power of graphs is the ability to understand a complex set of relationships as one holistic pattern: a path, a set of branches, a loop. TigerGraph’s pattern matching enhancement to TigerGraph’s GSQL query language makes it easier to do that.”
Copyright © 2018 RegTech Analyst