Charting the Future: Using Chart Question-Answering for Scalable Evaluation of LLM-Driven Data Visualizations

TL;DR

This paper introduces a VQA-based framework for scalable evaluation of LLM-generated data visualizations, highlighting current limitations and potential for rapid, automated assessment of chart quality.

Contribution

It presents a novel VQA-driven evaluation method for LLM-generated charts, enabling scalable and effective assessment of visual communication quality.

Findings

LLM-generated charts lag behind original charts in accuracy according to VQA metrics.

Few-shot prompting improves the quality of LLM-generated visualizations.

Significant progress is needed before LLMs can produce charts matching human quality.

Abstract

We propose a novel framework that leverages Visual Question Answering (VQA) models to automate the evaluation of LLM-generated data visualizations. Traditional evaluation methods often rely on human judgment, which is costly and unscalable, or focus solely on data accuracy, neglecting the effectiveness of visual communication. By employing VQA models, we assess data representation quality and the general communicative clarity of charts. Experiments were conducted using two leading VQA benchmark datasets, ChartQA and PlotQA, with visualizations generated by OpenAI's GPT-3.5 Turbo and Meta's Llama 3.1 70B-Instruct models. Our results indicate that LLM-generated charts do not match the accuracy of the original non-LLM-generated charts based on VQA performance measures. Moreover, while our results demonstrate that few-shot prompting significantly boosts the accuracy of chart generation, considerable progress remains to be made before LLMs can fully match the precision of human-generated graphs. This underscores the importance of our work, which expedites the research process by enabling rapid iteration without the need for human annotation, thus accelerating advancements in this field.