A Visual Approach for Health Information Exploration: Adaptive Levels of Visual Granularity and Interaction Analysis

TL;DR

The paper tackles the problem of static, non-adaptive health information by introducing apchis, a visual system that enables adaptive document exploration across multiple levels of detail. It combines established visualization techniques (Word Clouds, Tile Bars) with topic modeling and an interaction provenance framework to support personalized, non-linear information search in health materials, demonstrated via a formative study on a diabetes brochure. The main contributions include evidence of usability benefits over linear exploration, a provenance visualization approach to analyze and adapt user interactions, and a model for content/presentation adaptation to guide future development. This work advances health information literacy by enabling engaging, tailored consumption of complex medical content and lays groundwork for adaptive recommendations and bias mitigation in health information systems.

Abstract

The effective and targeted provision of health information to consumers, specifically tailored to their needs and preferences, is indispensable in healthcare. With access to appropriate health information and adequate understanding, consumers are more likely to make informed and healthy decisions, become more proficient in recognizing symptoms, and potentially experience improvements in the prevention or treatment of their medical conditions. Most of today's health information, however, is provided in the form of static documents. In this paper, we present a novel and innovative visual health information system based on adaptive document visualizations. Depending on the user's information needs and preferences, the system can display its content with document visualization techniques at different levels of detail, aggregation, and visual granularity. Users can navigate using content organization along sections or automatically computed topics, and choose abstractions from full texts to word clouds. Our first contribution is a formative user study which demonstrated that the implemented document visualizations offer several advantages over traditional forms of document exploration. Informed from that, we identified a number of crucial aspects for further system development. Our second contribution is the introduction of an interaction provenance visualization which allows users to inspect which content, in which representation, and in which order has been received. We show how this allows to analyze different document exploration and navigation patterns, useful for automatic adaptation and recommendation functions. We also define a baseline taxonomy for adapting the document presentations which can, in principle, be leveraged by the observed user patterns. The interaction provenance view, furthermore, allows users to reflect on their exploration and inform future usage of the system.

Figures (7)

• Figure 1: The proposed apchis supports dynamic levels of detail. A dl (top) allows users to select one particular document they want to explore. At document level, an interactive toc (bottom left) preserves the global linear structure of the document while a chapter's substructure and content is visualized with dedicated visualization techniques for textual and pictorial data.
• Figure 2: The main components of apchis shown by an example of exploring a German diabetes health brochure aok: toc, wc/hwc, is, tileb, topicb, snps, and fulltext. Different actions (illustrated as blue arrows) allow a user to navigate from one view to another.
• Figure 3: The evaluation regarding the prior knowledge and usefulness of apchis components by the cwt participants.
• Figure 4: The results of the forced choice evaluation between apchis and the PDF viewer w.r.t. the defined performance goals. The performance goals in information processing are sorted along a continuum from more abstract (left) to more specific goals (right).
• Figure 5: The adjacency matrix of transitions between the 'high-level' tools. The entries in the main diagonal corresponds to our notion of loops, while multiple edges populate the remaining cells. The cells reflecting transitions which are technically not possible are intentionally left blank.