Titulo Estágio
Alert similarity for investigation workflows
Áreas de especialidade
Engenharia de Software
Engenharia de Software
Local do Estágio
Coimbra
Enquadramento
When investigating an alert, analysts usually leverage multiple sources of data. A key source of data is the results of past investigations. Identifying relevant past cases in often non-trivial to do algorithmically, as there are many attributes for each investigation that may only partially match the current alert's attributes. We propose a machine learning approach to tackle this issue to assist in developing a similarity mapping between the current alert and past investigations.
Objetivo
When investigating an alert, analysts usually leverage multiple sources of data. A key source of data is the results of past investigations. Identifying relevant past cases in often non-trivial to do algorithmically, as there are many attributes for each investigation that may only partially match the current alert's attributes. We propose a machine learning approach to tackle this issue to assist in developing a similarity mapping between the current alert and past investigations.
Plano de Trabalhos - Semestre 1
"1. Onboard on risk strategy and operational processes at feedzai
2. Interview stakeholders (risk analysts) to understand better the operational flow
3. Review literature on existing approaches for similarity between documents
Expected results:
1. Literature review
2. Refine planning for the second semester based on review"
Plano de Trabalhos - Semestre 2
"Stages:
1. Define success criteria for similarity matching
2. Implement prototype of proposal algorithms for similarity matching
3. Refine approach based on results with stakeholders
4. Document approach and results
5. Draft final report and presentation"
Condições
Remunerated
Orientador
António Luís Correia
luis.correia@feedzai.com 📩