User-driven semantic classification for the analysis of abstract health and visualization tasks

Abstract

Present article outlines characteristics of a general task analysis in terms of digital health visualization evaluation and design. Furthermore, a number of methodological approaches are discussed. One example, in which a hierarchical structure was empirically built with semantic classification by 98 users, will be discussed together with the expected benefits of its successful implementation with respect to system development and human factors research on health data visualizations. It is concluded that experimental approaches to taxonomy construction offer considerable promise in capturing tasks which are relevant but that further investigation is needed validating and iteratively extending the abstract task structures. We thus recommend based on our experiences to conduct a combination of semantic classification with users and hierarchical task analysis to capture all needed task abstraction levels.