Evaluating personalised information retrieval: a perception of trust

Johnson, Frances (2019) Evaluating personalised information retrieval: a perception of trust. In: Second Annual Workshop on the Evaluation of Personalisation in Information Retrieval (WEPIR 2019), 09 March 2019 - 14 March 2019, Glasgow, Scotland.

Preview

Published Version Download (327kB) | Preview

Abstract

This position paper sets out a case for the evaluation of personalised information retrieval from the user perspective, and which focuses on the effect of personalisation on the searcher’s perception of a better information retrieval interaction. The proposed approach, complementing system performance evaluation, is about real users, real tasks and their interactions of query, search and assessment of the retrieved results, as well as their intentions driving the interaction. The components of user evaluation are discussed, and the searcher’s perception of usefulness is identified to specifically evaluate the effect of personalisation. The proposition is that a perception of ‘support to do a task’ is formed during use and which constitutes a major part of the user’s evaluation or judgment of trust. As such, it is suggested that, a perception of trust and its criteria provide a framework for evaluation. Further consideration is given to the construct of trust as an evaluative instrument with the goal of advancing personalisation in information interactions.