Titulo Estágio
Benchmarking graph analytics for fraud
Áreas de especialidade
Engenharia de Software
Engenharia de Software
Local do Estágio
Coimbra
Enquadramento
Financial fraud takes many forms. In particular, criminals behave in networks to try to ellude fraud detection systems. In this internship we propose a review of State-of-the-Art graph analytics to then benchmark them against real transactional data.
Objetivo
Financial fraud takes many forms. In particular, criminals behave in networks to try to ellude fraud detection systems. In this internship we propose a review of State-of-the-Art graph analytics to then benchmark them against real transactional data.
Plano de Trabalhos - Semestre 1
"Stages:
1. Review financial illicit activity mapping (scams, mule accounts, layering, etc)
2. Review State of the art Graph Analytics and Graph Mining techniques.
3. Define a methodology plan for benchmarking
4. Define success criteria/KPIs for the benchmarking
Expected results:
1. Documented methodology"
Plano de Trabalhos - Semestre 2
"Stages:
1. Implement the test-bed for all benchmarks.
2. Benchmark and iterate.
3. Documentation, final report, and presentation.
Expected results:
1. Documented benchmarking results
2. Applicable prototype of at least one graph analytics tool in a specific Feedzai case (if possible)"
Condições
Remunerated
Orientador
João Costa
joao.viriato@feedzai.com 📩