A three-agent workflow (processor, composer, validator) improves automatic data-to-chart generation on the VisEval benchmark, especially for multi-table queries.
Proceedings of the VLDB Endowment , 17(12):3920–3933
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
nvAgent: Automated Data Visualization from Natural Language via Collaborative Agent Workflow
A three-agent workflow (processor, composer, validator) improves automatic data-to-chart generation on the VisEval benchmark, especially for multi-table queries.