Interpreting the link value of similarity scores between illicit drug specimens through a dual approach, featuring deterministic and Bayesian frameworks.

Illicit drug trafficking and in particular amphetamine-type stimulants continue to be a major problem in Australia. With the constant evolution of illicit drugs markets, it is necessary to gain as much knowledge about them to disrupt or reduce their impact. Illicit drug specimens can be analysed to generate forensic intelligence and understand criminal activities. Part of this analysis involves the evaluation of similarity scores between illicit drug profiles to interpret the link value. Most studies utilise one of two prominent score evaluation approaches, i.e. deterministic or Bayesian. In previous work, the notion of a dual approach was suggested, which emphasised the complementary nature of the two mentioned approaches. The aim of this study was to assess the operational capability of a dual approach in evaluating similarity scores between illicit drug profiles. Utilising a practical example, link values were generated individually from both approaches, then compared in parallel. As a result, it was possible to generate more informed hypotheses, relating to specimen linkage, due to the greater wealth of information available from the two approaches working concurrently. Additionally, it was shown that applying only one approach led to less information being generated during analysis as well as potentially important links between illicit drug specimens being missed.