Interactive system links entities and documents handling uncertainty
EntityWeaver: Visual Exploration and Curation of Named-Entity Relationships in Document Collections
Human-Computer Interaction
Summary
Collections of documents often mention people, places, or things called named entities, but it's hard to track how these mentions relate to real-world entities due to uncertainties. The authors created a tool called EntityWeaver that visualizes these connections as a graph, showing how strong or uncertain each link is. Users can adjust how uncertainty appears, explore large collections efficiently, and jump directly into full text to verify or fix confusing parts. This helps uncover patterns in documents while letting people clean up and understand ambiguous or unclear links.
What this means in practice
- •For digital archivists: Clean and explore large text archives by interacting with uncertain entity connections and source documents to improve data quality.
- •For media analysts: Analyze news or report collections for patterns in named entities while managing ambiguous references through direct text inspection.
Authors
Uroš Šmajdek, Ciril Bohak
Abstract
We present an interactive visualization system for exploring named entities and their relationships across document collections, with a strong focus on handling uncertainty and supporting both distant and close reading. The system is built around a graph that links documents, entity mentions, and entities. Uncertainty from mention-to-entity linking is included directly in this graph, so users can see where connections are strong, weak, or ambiguous. A transfer-function control, inspired by approaches in scientific visualization, allows users to adjust how this uncertainty is displayed, making it easy to tune the visualization for different datasets and research questions. The system also provides direct access to the full source texts in a coordinated view, enabling quick context checks, resolving ambiguous cases, and correcting digitization errors. Exploration is further supported through a multi-stage filter query builder, mini-map navigation for large graphs, and export options for downstream analysis. By combining uncertainty-aware graph visualization with direct interaction in the source texts, the system provides a unified workflow that supports both large-scale pattern discovery and curation of named-entity-based document collections. The design choices were supported by the domain experts, who were also involved in the initial system evaluation.