Adoption of graph database technology by enterprises has been steadily rising in the last ten years. Neo4j, the popular graph database vendor founded in 2007, claims that almost half of Forbes Global 2000 companies use some form of its graph database.
Compared to traditional relational databases that store data in highly structured environments with predetermined columns, graph databases store data and relationships in a graph made up of nodes and connected via relationships. The power of graph database is that it allows users to discover rich insights from connections within their data set, rather than within a constrained environment where one entry corresponds to a set of columns in a table which makes connections between entries elusive.
Use cases for graph database are far-extending and ever growing. Banks use graph database to detect complex patterns of transactions that could be fraudulent. Retailers pull together customer’s purchase history, inventory data and social sentiment data in a graph database to develop real-time recommendation engines. Increasingly, there is also a compelling case of using graph database within the risk and compliance world.
The flagship example of how graph database was applied to the risk and compliance world is the now-widely known International Consortium of Investigative Journalists (ICIJ) Offshore leaks database. The database showed how graph technology was useful in illuminating complex data sets during an investigation. Intelligent queries such as ‘How was the president of Azerbaijian connected to offshore accounts?’ seen here could be efficiently run on a graph database. The query returns a trail of how Azerbaijani President Ilham Aliyev and his family members were named as beneficiaries of offshore companies in the British Virgin Islands.
When conducting compliance checks on a subject, practitioners encounter and process a multitude of different data sets, such as corporate records, Politically-Exposed-Persons (PEP) lists, sanctions lists, adverse media articles and litigation records. As it is, practitioners today either manually or with aid of visualization tools weave a relationship map of all the artefacts that are relevant to the scope of the investigation themselves. Leveraging on graph database is an intelligent extension of this process.
Graph database allow for fast relationship-based queries so users could get answers to risk-based queries such as “Is the subject in litigation with any of his suppliers or family members?”, “Is the subject connected to any politically exposed persons (PEP)s in five degrees of separation?” and “Who are the ultimate beneficial owners of the company?” In this way, practitioners can leverage on the data you’ve already gathered and processed without manually making the links themselves.
Datarama leverages on intelligent graph database algorithms to make compliance and due diligence checks more efficient. Find out how we do it at datarama.com, or contact us at firstname.lastname@example.org.