Topic Similarity Networks: Visual Analytics for Large Document Sets

TL;DR

This paper presents a visual analytics approach using topic similarity networks to enhance the interpretability of LDA models, enabling better exploration and understanding of large text collections through effective visualization and analysis.

Contribution

It introduces efficient methods for constructing and labeling topic similarity networks, improving the interpretability and exploration of large-scale topic models.

Findings

Networks reveal non-obvious document connections

Visualizations help characterize large text collections

Case studies demonstrate practical utility

Abstract

We investigate ways in which to improve the interpretability of LDA topic models by better analyzing and visualizing their outputs. We focus on examining what we refer to as topic similarity networks: graphs in which nodes represent latent topics in text collections and links represent similarity among topics. We describe efficient and effective approaches to both building and labeling such networks. Visualizations of topic models based on these networks are shown to be a powerful means of exploring, characterizing, and summarizing large collections of unstructured text documents. They help to "tease out" non-obvious connections among different sets of documents and provide insights into how topics form larger themes. We demonstrate the efficacy and practicality of these approaches through two case studies: 1) NSF grants for basic research spanning a 14 year period and 2) the entire English portion of Wikipedia.