Novel Visualisation of Financial Data

Industry Challenge

Fraud prevention requires remote and distributed teams to interact with large quantities of complex, multi-dimensional data which is time-consuming, costly, and prone to error. The exercise often requires teams to co-locate to ensure effective collaboration and secure data management.

Background:

This industry partner was interested in showcasing their engagement with Academia. By inventing a new way of interacting with large quantities of data, they aim to use VR to disclose suspicious financial transactions.

Challenges:

– Quickly filter through large quantities of data
– Visualise connections between people and organisations
– Highlight suspicious content based on real-time parsing of relevant news stories
– Let investigators collaborate in the VR space

ADAPT Solution

Input: Real Time stream of financial transaction data connected to a database of people and organisations together with meta-information around types of transactions, frequency, categories, and geographies.

Technology: Unity was used to create a 3D world in which data was represented as physical entities that could be filtered in several ways. Natural Language Processing (NLP) is employed to understand news articles and connect events to specific data points.

Output: The application allows financial fraud investigators to collaborate in real-time within the same VR environment. This can be done without the need to be in the same physical location.

Industry Benefits and use cases

• Visual representation of large datasets.
• Relevant filtering options.
• Secure access and presentation.
• Endless space in VR.
• Fraud investigation: Monitor people on suspicious activities.
• Social media analysis to determine how people or organisations are connected with each other.