Use of an Automated System to Evaluate Feature Dissimilarities in Handwriting Under a Two‐Stage Evaluative Process*†

Abstract

The two-stage evaluative process is an established framework utilized by forensic document examiners (FDEs) for reaching a conclusion about the source(s) of handwritten evidence. In the second, or discrimination, stage, the examiner attempts to estimate the rarity of observations in a relevant background population. Unfortunately, control samples from a relevant background population are often unavailable, leaving the FDE to reach this determination based on subjective experience. Automated handwriting feature recognition systems are capable of performing both feature comparison and discrimination, yet these systems have not been subjected to empirical validation studies. In the present study, we repurposed a commercially available automated system to generate empirical distributions for ranking feature dissimilarity scores among pairs of handwritten phrases. The blinded results of this automated process were used to survey an international cohort of 36 FDEs regarding their strength of support for same-and different-writer propositions. The survey served to cross-validate FDE decision-making under the two-stage approach. Results from the survey demonstrated a clear pattern of response consistent with ground truth. Predictive regression analyses indicated that the automated feature dissimilarity scores and the log of their cumulative distribution functions accounted for 72% of the variability in FDE opinions. This study demonstrated that feature dissimilarity scores acquired using automated processes and their distributions are closely aligned with FDE decision-making processes supporting the heuristic value of the two-stage evaluative framework.

Elucidating the relationships between two automated handwriting feature quantification systems for multiple pairwise comparisons

“…The study recruited 33 volunteer writers from the San Diego Sheriff's Crime Laboratory; each subject was asked to write six phrases from the London Letter [15] and to repeat each phrase five times using both handprinting and cursive writing styles (for a total of 60 writing samples per subject). Handwriting data from these subjects were used in two prior studies aimed at further understanding the decision‐making process of forensic document examiners [16,17]. Subjects were asked to write the handwriting sample phrases with an inking pen on lined papers placed on top of a Wacom (Intuos Pro, model PTH‐660) digitizing tablet.…”

“…For this study, we modified the scoring output (but not the feature extraction) of FLASH ID ® , as previously described in Fuglsby et al [16]. The output of FLASH ID ® encodes all the graphemes in a document relative to a reference set of writers (in this case, 50 writers from the “FBI100” data set, described in Saunders et al [18]; the reference set is a term used in FLASH ID ® to denote a list of possible writers of interest for the original recommendation system).…”

“…For this study, we modified the scoring output (but not the feature extraction) of FLASH ID ® , as previously described in Fuglsby et al [16].…”

Interpol questioned documents review 2019–2022

“…Fuglsby et al [ 134 ], conducted an experiment to deploy an automated feature extraction program to generate feature dissimilarity scores and population distribution functions for ranking these feature dissimilarity scores among pairs of handwritten phrases across different phrases and styles of handwriting. A second experiment was to utilize these dissimilarity scores and distribution functions to design a series of difficult case scenarios for FDEs to evaluate.…”

Application of nonparametric quantifiers for online handwritten signature verification: A statistical learning approach

“…The significance of handwritten signatures extends into diverse applications, encompassing online banking, credit cards, and cheque processing mechanisms [2]. Moreover, biometric systems play a pivotal role in the authentication and validation of passports, employing methods such as signature verification [3,4] and user behavioral characteristic verification in the realm of digital forensics [5][6][7]. Signing is a minimally invasive form of identification compared to methods like DNA, fingerprint, and blood analysis.…”