Are Taylor's posts risky? Evaluating cumulative revelations in online personal data

Azzopardi, Leif, Briggs, Jo ORCID: https://orcid.org/0000-0002-4041-1918, Duheric, Melissa, Nash, Callum, Nicol, Emma, Moncur, Wendy and Schafer, Burkhard (2022) Are Taylor's posts risky? Evaluating cumulative revelations in online personal data. In: SIGIR '22: The 45th International ACM SIGIR Conference on Research and Development in Information Retrieval, 11 July 2022 - 15 July 2022, Madrid, Spain.

Preview

Accepted Version Available under License In Copyright. Download (5MB) | Preview

Abstract

Searching for people online is a common search task that most of us have performed at some point or other. With so much information about people available online it is often amazing what one can find out about someone else -- especially when information taken from different sources is pieced together to create a more detailed picture of the individual, and then used to make inferences about them (leading to cumulative revelations ). As such, the relevance of one piece of information is often conditional and dependent on other pieces of information found. This creates interesting and novel challenges in evaluating informationrelevance when searching personal profiles, posts and related information about an individual, as well as the potential risks that can arise from such revelations. In this demonstration paper, we present a tool designed to investigate how people assess and judge the relevance and potential risks ofsmall, apparently innocuous pieces of information associated with fictitious personas, such as Taylor Addison, when searching and browsing online profiles and social media. The demonstrator also comprises a cyber-safety tool, which aims to provide education and raise awareness of the potential risks of cumulative revelations. It does so by engaging participants in different scenarios where the relevance of individual information items depends on the searcher and their particular underlying motivation.