Improving Knowledge Graph Understanding with Contextual Views -- Extended

TL;DR

This paper introduces the InK Browser, a tool that leverages ontological information in knowledge graphs to enhance navigation, visualization, and discovery through modular, interactive views, improving user experience.

Contribution

The paper presents the InK Browser, a novel tool that utilizes ontological data to provide flexible, modular perspectives for better knowledge graph exploration.

Findings

User survey indicates improved KG exploration

Modular views facilitate better understanding of KGs

Tool enhances navigation and visualization capabilities

Abstract

Navigating, visualizing, and discovery in graph data is frequently a difficult prospect. This is especially true for knowledge graphs (KGs), due to high number of possible labeled connections to other data. However, KGs are frequently equipped with an ontology as a schema. That is, it informs how the relationships between data may be constrained. This additional information can be leveraged to improve how (knowledge) graph data can be navigated, visualized, or otherwise utilized in a discovery process. In this manuscript, we introduce the Interactive Knowledge (InK) Browser. This tool specifically takes advantage ontological information (i.e., knowledge) when found in KGs. Specifically, we use modular views that provide various perspectives over the graph, including an interactive schema view, data listings based on type, neighborhood connections, and geospatial depiction (where appropriate). For this manuscript, we have evaluated the basic premise of this tool over a user group ($n= With this grown user survey, we continue to evaluate how scalable tools, including flexible views, can make KG exploration easier for a range of applications.)