DyNamiTE - Digital fightiNg Tax Evasion

The project, aimed at studying innovative techniques for identyfying potential tax evaders, is funded by the Fondazione Cassa di Risparmio di Cesena, and is carried out on a real case study provided by Comune di Cesena in collaboration with Onit Group.

The project started from the consideration that Municipalities, and more in general, public Administrations have nowadays access to a huge amount of information: internally-managed ones, open data, and data managed by other public bodies. Unfortunately such data are typically “dirty” and poorly integrated, while the full exploitation of such data could impact on the effectiveness of the social strategies thus enabling several smart applications.

Our activity goes in two directions:

• Database integration: we coupled standard database integration technique with smart cleansing and approximate joins one. This allowed to increase up to 20% the percentage of unmatching records.
• Smart exploitation of integrated and cleansed data: we prototyped two different applications in the area of tax analysis:
  • Tax evaders identification: we applied data mining techniques to the integrated data to identify evasion pattern in the area of real estates and undeclared rents. Short lists of potential evaders returned by our techniques have been verified by the tax department of the municipality that started the tax assessment process.
  • Fiscal Equity simulation: we created a what-if prototype simulating the impact of local taxes on different types of families (e.g., retired couples, young couples, single, couples with 1,2,3 children). The goal of the application is to analyze the variation of the distribution of taxes determined by changes in tax parameters (e.g., tax shares, exemption thresholds). The integrated data are necessary to estimates the average wealth of different types of families and the impact of taxes.