Using case specific experiments to evaluate fingermarks on knives given activity level propositions.

Area of Science:

• Forensic Science
• Probability Theory
• Computational Intelligence

Background:

• Bayesian networks are valuable for evaluating forensic evidence at the activity level.
• Assessing fingermarks requires robust methods to link evidence to specific actions.

Purpose of the Study:

• To demonstrate using case-specific experiments to assign probabilities for Bayesian network nodes.
• To evaluate fingermarks in the context of activity level propositions.

Main Methods:

• Conducted experiments on fingermark transfer, persistence, and recovery on knives.
• Modeled two Bayesian networks using experimental data to assess different probability assignments.
• Applied Bayesian networks to evaluate fingermark evidence in disputed activities.

Main Results:

• Experimental data successfully informed probability assignments within the Bayesian networks.
• The study explored how different uses of experimental data impact network evaluation.
• Bayesian network analysis showed the potential of fingermarks for activity level propositions.

Conclusions:

• Case-specific experiments are effective for quantifying probabilities in Bayesian network models for fingermark analysis.
• Bayesian networks provide a framework for evaluating the evidential weight of fingermarks concerning disputed activities.
• The methodology holds potential for broader application in forensic science for activity level assessments.