Likelihood ratio data to report the validation of a forensic fingerprint evaluation method

Area of Science:

• Forensic Science
• Biometrics
• Pattern Recognition

Background:

• Likelihood ratios (LR) are essential for evaluating forensic evidence.
• Validating LR methods requires robust datasets for reliable forensic conclusions.
• Existing datasets may not fully support the validation of LR methods in forensic casework.

Purpose of the Study:

• To provide shared likelihood ratio (LR) data derived from minutiae comparisons of fingermarks and fingerprints.
• To facilitate the validation of LR methods used in forensic evidence evaluation.
• To serve as a baseline for reproducibility and training in generating forensic validation reports.

Main Methods:

• LR values were computed by comparing 5-12 minutiae from fingermarks against full fingerprints.
• A validation procedure was established for LR methods used in forensic evidence evaluation.
• Real forensic fingerprints were used for validation, while simulated data were used for development.

Main Results:

• The study presents shared LR data crucial for the validation of forensic evidence evaluation methods.
• The data enables reproducibility of validation experiments and training for validation report generation.
• Justification and motivation for the chosen methods and validation criteria are provided.

Conclusions:

• The shared LR data are a necessary asset for validating LR methods in forensic science.
• These data support the development of standardized validation reports for forensic techniques.
• While original images are withheld for privacy, the core LR data are available for scientific scrutiny.