Co-Designing Unstructured Text Data Visualization Systems

TL;DR

This work addresses the lack of generalizable qualitative visualization tools for unstructured text by proposing a co-design approach using focus groups to elicit requirements for a flexible dashboard that can reveal entities, actors, and multiple narratives. The authors contrast existing tools (e.g., ConceptVector, Storifier, TopicDrivers) and their limitations in supporting multi-perspective storytelling and short-text analysis, motivating a generalized, customizable solution. Guided by Kerzner's Creative Visualization framework and Munzner's nine-stage design study, they outline a methodology for gathering requirements, characterizing problems, and developing low-fidelity prototypes to explore a broad solution space. The anticipated contribution is a transferable design process and a configurable visualization tool capable of supporting exploratory analysis of narratives and relationships in large unstructured text corpora, with practical implications for researchers and practitioners analyzing complex qualitative data.

Abstract

We present our in-progress work on co-designing a visualization tool for presenting unstructured text. We have conducted a focus group with a variety of professionals who regularly analyze large corpora of unstructured text. Our preliminary insights indicate there is an unmet need to visually explore the dynamics between entities and actors extracted from unstructured text. Additionally, large corpora contain multiple perspectives on the same series of events. There is a need to disentangle these perspectives and visually show the multiple narratives present in the data. In our future work, we will co-design low-fidelity prototypes to create a broad consideration space of possible solutions for visualizing unstructured text.