The client’s challenge
The DWSE (Department of Employment and Social Economy) – part of the Flemish government – is responsible for policy on employment and social economy. DWSE also distributes both electronic and paper service vouchers that can be used as subsidized payment for one hour of household help. During manual audits, several violations against the specific rules of using service vouchers were found. Our member Vectr.consulting helped set up an AI application to help them detect fraud more easily and efficiently.
Flemish government
De Cronos Groep has been working alongside the Flemish Government for quite some time now. We are working with different sections of the organisation, like the Department of Employment and Social Economy and have plenty of AWS workloads actively running there.
Our solution
To increase the fraud detection rate, an AI-based application has been built using AWS, combining techniques such as outlier detection, time-series analysis and graph analytics with machine learning, resulting in AI models that are capable of detecting patterns indicative of fraud.
Workflow: The data is initially analyzed using standard outlier-detection techniques in regression models. Having 11 known fraud cases within the data set allowed us to test model hypotheses. This yielded good models for most of the known fraud cases, but some remained under the radar.
We extended the research into the time and connection domains. We built time series to find behavioural patterns in time, and outliers thereof. This allowed a refinement of the detection model, but still missed two important known frauds.
The final step was to connect all data and build an interaction graph of all connected entities in the data model. This was done using a specialized graph database and apply graph analytics on it.
Results
The result was astonishing: not only the upfront known fraud cases were clearly picked up, several other candidate fraud cases were detected. The application offers the inspectors a useful tool tot target investigations. After the initial increase in the fraud detection rate, the AI tool now successfully helps preventing fraud.