Visualizing Dimension Coverage to Support Exploratory Analysis

Abstract

Data analysis involves constantly formulating and testing new hypotheses and questions about data. When dealing with a new dataset, especially one with many dimensions, it can be cumbersome for the analyst to clearly remember which aspects of the data have been investigated (i.e., visually examined for patterns, trends, outliers etc.) and which combinations have not. Yet this information is critical to help the analyst formulate new questions that they have not already answered. We observe that for tabular data, questions are typically comprised of varying combinations of data dimensions (e.g., what are the trends of Sales and Profit for different Regions?). We propose representing analysis history from the angle of dimension coverage (i.e., which data dimensions have been investigated and in which combinations). We use scented widgets [30] to incorporate dimension coverage of the analysts' past work into interaction widgets of a visualization tool. We demonstrate how this approach can assist analysts with the question formation process. Our approach extends the concept of scented widgets to reveal aspects of one's own analysis history, and offers a different perspective on one's past work than typical visualization history tools. Results of our empirical study showed that participants with access to embedded dimension coverage information relied on this information when formulating questions, asked more questions about the data, generated more top-level findings, and showed greater breadth of their analysis without sacrificing depth.

Characterizing Exploratory Visual Analysis: A Literature Review and Evaluation of Analytic Provenance in Tableau

“…Behavior Graphs : Problem‐solving behavior can be modeled as a set of states along with a set of operations for moving between states [New72], including visual analysis [WMA∗16b, STM17, WQM∗17, SvW08, ST15, jJKMG07]. Graphs can capture more complex paths and patterns, such as back‐tracking or state revisitation.…”

“…Wongsuphasawat et al define depth and breadth in terms of the total unique states analyzed [WMA∗16b, WQM∗17] (also considered by Sarvghad et al [STM17]). However the interactions that produce these states, and thus the states themselves, are not independent.…”

Supporting Handoff in Asynchronous Collaborative Sensemaking Using Knowledge-Transfer Graphs

“…Coordination and synthesis strategies between asynchronous data analysts were also previously studied [37,40,41], sharing our interest in understanding and consolidating work of previous investigators. Robinson [37] focused on the co-located synthesis of findings after an asynchronous distributed analysis phase.…”

“…Robinson [37] focused on the co-located synthesis of findings after an asynchronous distributed analysis phase. Sarvghad et al [40,41] found visualizing data dimension coverage of a previous analyst's exploration can promote awareness and understanding, as well as the question formation process, in tabular data analysis.…”

How Domain Experts Structure Their Exploratory Data Analysis: Towards a Machine-Learned Storyline

“…In fact, we seem to go faster than we have time to note down" (a participant from study [5]). Moreover, our domain experts were not able to build a clear mental model of past exploration, nor did they have any means to evaluate whether they had a robust or exhaustive exploration strategy [26]. Indeed, the lack of support tools for sensemaking of past exploration is common in many EDA systems [19,10,16].…”