Unexplored Frontiers: A Review of Empirical Studies of Exploratory Search

Abstract

This article reviews how empirical research of exploratory search is conducted. We investigated aspects of interdisciplinarity, study settings and evaluation methodologies from a systematically selected sample of 231 publications, including a total of 172 articles with empirical studies. Our results show that exploratory search is highly interdisciplinary, with the most frequently occurring publication venues including high impact venues in information science, information systems and human-computer interaction. However, taken in aggregate, the breadth of study settings investigated was limited. We found that a majority of studies (77%) focused on evaluating novel retrieval systems as opposed to investigating users' search processes. Furthermore, a disproportionate number of studies were based on scientific literature search (20.7%), a majority of which only considered searching for Computer Science articles. Study participants were generally from convenience samples, with 75% of studies composed exclusively of students and other academics. The methodologies used for evaluation were mostly quantitative, but lacked consistency between studies and validated questionnaires were rarely used. In discussion, we offer a critical analysis of our findings and suggest potential improvements for future exploratory search studies.