Interactive Search Profiles as a Design Tool

摘要

Interactive Information Retrieval (IIR) research studies how users interact with IR systems and evaluates the users' satisfaction with the retrieval process. Thus, it focuses on how users behave when they interact with IR systems. The involvement of potential users, and access to dynamic and individual information needs are essential elements in IIR studies. User-oriented evaluations investigate the ability of users to engage with a system, thus requiring real users to be involved in the evaluation process. Despite its importance in driving the design of better and more usable IR systems, the user-oriented evaluation approach has been criticised for: (ⅰ) being expensive and time consuming in its application, (ⅱ) failing to capture a holistic understanding of the context, as it focuses on users, (ⅲ) delivering not reusable data, often of qualitative nature, and (ⅳ) generating experiments and results that are not reproducible. In this paper, we propose an alternative approach, based on data extracted from Log Files (LF) and automatically extracted Usage Pattern (UP)s. Interactive Search Profiles (ISP)s are then created as tools to guide the design of more usable IIR systems, specifically considering Digital Libraries (DL).