Authorship attribution studies use features to analyse an author’s idiolect. The current study investigates the usage of some of these features over time. Over 120,000 WhatsApp text messages of...Show moreAuthorship attribution studies use features to analyse an author’s idiolect. The current study investigates the usage of some of these features over time. Over 120,000 WhatsApp text messages of seven Dutch participants were processed and analysed. Using relative frequencies of specific features, distributions of authors’ usage of the features throughout a period of three to five years was extracted and visualized to determine which features are consistent over time and which are not. The goal of the study was to determine a ranking of the features, which could speak to their reliability in authorship attribution, especially when there is a considerable time interval between samples. The results suggest that common abbreviations, filler words and message length are largely consistent features, while emojis are largely inconsistent.Show less
