Forensic intelligence on illicit markets: The example of watch counterfeiting

Population identification strategies for counterfeit coin detection

Advances in manufacturing, 3-d imaging, and globalization have led to a rise in fraudulent coinage and a world-wide interest in coin authentication. Modern manufacturing methods allow the alloy, construction, and struck image of coins to be more readily reproduced. Larger coin denominations and efforts to reduce the cost of coining add additional incentive. Detection of fraudulent coinage can parallel authentication of food, beverages, and manufactured goods by studying product-inherent features. Reverse-quality-engineering provides clues to authenticity. One promising method is in the use of finite mixture models to compare individual measurements of groups of coins to assist in authentication. An example is provided using the coin weights of two groups of coins. Authentication of a Questioned set of coins is explored, comparing the weight population of Example coins drawn from circulation with the weights of a Questioned set drawn from an unknown origin. In the test, just over half of the Questioned coin set matched the distribution of the Example coin set. The other portion, nearly half of the coin sample, did not match the Example coins drawn from circulation. If this were combined with a similar analysis of other coin properties, similar results would help validate the finding. The example shows that groups of coins can be authenticated by using one or more measures of properties of populations of Questioned coins versus Example coins that are largely authentic.

Determining the impact of unknown individuals in criminality using network analysis of DNA matches

Criminal offenders missing from police files limit the capacity to reconstruct criminal networks for criminological research and operational purposes. Recent studies show that forensic DNA databanks offer potential to address this problem, through large-scale analysis of DNA matches, many of which involve unidentified offenders. Applying social network analysis (SNA) to 18 years of DNA match data from Québec, Canada, we found that 1400 unknowns do not occupy more marginal positions in the network than 13,000 known offenders, and explain up to 18% of SNA values (e.g., betweenness centrality) for the latter while supporting 46% of their clustering values. Our results contrast with previous studies, showing moreover that unknown individuals who are positioned centrally in a network may have a larger impact than previously expected on investigation policing with implications for forensic intelligence.