Visual Exploration of Relationships and Structure in Low-Dimensional Embeddings.

Klaus Eckelt,Andreas Hinterreiter,Conny Walchshofer,Vaishali Dhanoa,Patrick Adelberger,Marc Streit,Moritz Heckmann,Christina Humer,Christian Steinparz

doi:10.1109/tvcg.2022.3156760

Klaus Eckelt, Andreas Hinterreiter + Show 7 more

Open Access

https://doi.org/10.1109/tvcg.2022.3156760

Copy DOI

Abstract

In this work, we propose an interactive visual approach for the exploration and formation of structural relationships in embeddings of high-dimensional data. These structural relationships, such as item sequences, associations of items with groups, and hierarchies between groups of items, are defining properties of many real-world datasets. Nevertheless, most existing methods for the visual exploration of embeddings treat these structures as second-class citizens or do not take them into account at all. In our proposed analysis workflow, users explore enriched scatterplots of the embedding, in which relationships between items and/or groups are visually highlighted. The original high-dimensional data for single items, groups of items, or differences between connected items and groups is accessible through additional summary visualizations. We carefully tailored these summary and difference visualizations to the various data types and semantic contexts. During their exploratory analysis, users can externalize their insights by setting up additional groups and relationships between items and/or groups. We demonstrate the utility and potential impact of our approach by means of two use cases and multiple examples from various domains.

Highlights

The challenge of making high-dimensional data accessible for visualizations in two-dimensional space is typically addressed by dimensionality reduction (DR)
We demonstrate our visual exploration approach by means of two use cases from different domains: (1) the analysis of openings in professional chess games and (2) the analysis of cancer patient cohorts based on genomics data
We presented an interactive visual exploration approach for structural relationships in embeddings of high-dimensional data

Summary

Introduction

The challenge of making high-dimensional data accessible for visualizations in two-dimensional space is typically addressed by dimensionality reduction (DR). A plethora of powerful visualization and interaction methods has been proposed for interpreting and exploring scatterplots of dimensionally reduced data, referred to as embeddings [1]. Most existing approaches do not take a defining characteristic of many real-world datasets into account: structural relationships between items and groups of items. Item-to-item relationships, for instance, can result from an inherent (temporal) ordering of data items. Item-to-group associations can be based on shared categorical values or user-defined group labels. Group-to-group relationships are defining properties in hierarchical datasets

Objectives

Findings

Discussion

Conclusion