A likelihood ratio-based forensic text comparison in predatory chatlog messages

Abstract

An experiment in Forensic Text Comparison (FTC) within the Likelihood Ratio (LR) framework is described, which determines the strength of authorship attribution evidence from chatlog messages using so-called lexical features. More specifically, in this study I will investigate 1) the degree of evidential strength (or LR) that can be obtained from chatlog messages and 2) how the performance of the FTC system and the magnitudes of the LRs are influenced by the sample size for modelling. The performance of the system is assessed using the log-LR cost (Cllr) and the magnitudes of the obtained LRs are visually presented as Tippett plots. It is demonstrated in this study that you can use the lexical features within the LR framework to discriminate same-author and different-author chatlog messages.