Visual Text Analytics using Semantic Networks and Interactive 3D Visualization

Abstract

Facing the growth of textual information, the analysis of unstructured text data remains a challenge for visual analytics. Most text visualizations are based on models that use word frequencies for text vectorization and representation. As semantics reveal from word relations, we propose an integrated visual text analytics approach that utilizes semantic networks in an interactive 3D workspace for exploration and analysis. Semantic networks act as an intermediary structure for data modeling and interactive visualization to support the visual analytics process. Focussing on the integration of text analysis and visualization, this paper describes our system design and its preliminary implementation. Discussing typical usage scenarios and a practical field-test, we present a strategy for text exploration and analysis to illustrate the usage of our system in an exemplary use case.