Envisage: Towards Expressive Visual Graph Querying

TL;DR

Envisage is an interactive visual graph querying system that enhances expressiveness and usability for complex graph queries, enabling users to construct, verify, execute, and analyze queries without extensive coding.

Contribution

It introduces a novel four-stage system that supports intuitive query construction, verification, progressive execution, and result analysis for complex graph querying.

Findings

Effective in constructing complex queries

Validated through case studies and user interviews

Improves usability for graph analysts

Abstract

Graph querying is the process of retrieving information from graph data using specialized languages (e.g., Cypher), often requiring programming expertise. Visual Graph Querying (VGQ) streamlines this process by enabling users to construct and execute queries via an interactive interface without resorting to complex coding. However, current VGQ tools only allow users to construct simple and specific query graphs, limiting users' ability to interactively express their query intent, especially for underspecified query intent. To address these limitations, we propose Envisage, an interactive visual graph querying system to enhance the expressiveness of VGQ in complex query scenarios by supporting intuitive graph structure construction and flexible parameterized rule specification. Specifically, Envisage comprises four stages: Query Expression allows users to interactively construct graph queries through intuitive operations; Query Verification enables the validation of constructed queries via rule verification and query instantiation; Progressive Query Execution can progressively execute queries to ensure meaningful querying results; and Result Analysis facilitates result exploration and interpretation. To evaluate Envisage, we conducted two case studies and in-depth user interviews with 14 graph analysts. The results demonstrate its effectiveness and usability in constructing, verifying, and executing complex graph queries.