Organization responsible: SNN Adaptive Intelligence, Nederlands Forensisch Instituut
- Bert Kappen (project leader, researcher, SNN)
- Wim Wiegerinck (researcher, SNN)
- Willem Burgers (researcher/programmer, SNN)
- Ender Akay (programmer, SNN)
- Kees Albers (researcher, SNN)
- Martijn Leisink (researcher, SNN)
- Carla Bruijning (researcher, NFI)
The aim of the research and knowledge transfer project is to improve forensic institute’s missing person screening and matching routine for victim identification based on DNA profiles, e.g. in crisis situations using Bayesian network modeling and inference methods.
The aim of the research and knowledge transfer project is to improve the forensic institute’s missing person screening and matching routine for victim identification based on DNA profiles, e.g. in crisis situations. The research assumption is that Bayesian network modeling and inference methods can greatly contribute in this aim. The specific research and knowledge transfer question is then how to apply Bayesian network methods to this problem. This research is in collaboration with NFI (Netherlands Forensic Institute).
How can Bayesian networks be modeled in this specific application field, including all the desiderata (see appendix I Bonaparte proposal), such that the tool can handle many data automatically without putting too much modeling burden on the forensic researcher.
We investigate the forensic domain knowledge (literature, NFI experts). We define a model. We implemented a generic Bayesian network prototype model in Matlab, study its properties – and optimize where possible, validate this implementation with NFI experts. After passing this test, we implement a JAVA computation core using BayesBuilder to demonstrate the real-world viability of the approach. Results of the JAVA implementation will be validated by the Matlab implementation. Further extensions will be investigated and demonstrated in the same way.