One of the most cost-effective activities Italian town halls are supposed to carry out is tax evaders detection. That is why the city of Florence asked Evodevo to develop the Florence Fraud Detection System, a Decision Support System to help municipalities employees in automating the detection process.
The Florence Fraud Detection System consists of several databases merging into an Oracle 12c db with Spatial & Graph option with a description logic rules layer on top. The rules were derived from a study about the tax fraud domain in the specific section of Italian citizens resident abroad, in order to classify the citizens of Florence into different levels of suspiciousness. The system includes a Joseki SPARQL endpoint customization with pre-written semantic queries.
The session will consist of: an introduction to the context; a technical overview; a system demonstration.