2021/02/14 by Doris Jung‐Lin Lee, Vidya Setlur, Lee, Doris Jung-Lin +7
Computer Science · #Data Visualization and Analytics #Data Analysis with R #Data Stream Mining Techniques
paper · pdf · doi:10.48550/arxiv.2102.07070
Visualization recommendation (VisRec) systems provide users with suggestions\nfor potentially interesting and useful next steps during exploratory data\nanalysis. These recommendations are typically organized into categories based\non their analytical actions, i.e., operations employed to transition from the\ncurrent exploration state to a recommended visualization. However, despite the\nemergence of a plethora of VisRec systems in recent work, the utility of the\ncategories employed by these systems in analytical workflows has not been\nsystematically investigated. Our paper explores the efficacy of recommendation\ncategories by formalizing a taxonomy of common categories and developing a\nsystem, Frontier, that implements these categories. Using Frontier, we evaluate\nworkflow strategies adopted by users and how categories influence those\nstrategies. Participants found recommendations that add attributes to enhance\nthe current visualization and recommendations that filter to sub-populations to\nbe comparatively most useful during data exploration. Our findings pave the way\nfor next-generation VisRec systems that are adaptive and personalized via\ncarefully chosen, effective recommendation categories.\n