ChartOptimiser: Task-driven Optimisation of Chart Designs

2025-04-14

AI Summary

Manual chart optimization is inefficient due to the vast design space and strong parameter coupling, especially for task-specific visualization needs. Method: This paper proposes an analysis-task-driven automatic chart optimization framework supporting four canonical tasks—extreme value detection, numeric retrieval, pairwise comparison, and derived computation. It introduces a novel eight-dimensional perceptual objective function integrating visual salience, textual legibility, color preference, whitespace ratio, and other human-centered metrics. For the first time, it enables end-to-end optimization driven by user information needs and contextual constraints, combining Bayesian optimization with multi-objective perceptual modeling. Contribution/Results: Human-subject experiments and task-performance evaluations across 12 bar charts and four task types demonstrate statistically significant improvements (p < 0.01) in task completion rate, chart clarity, and visual aesthetics. The framework further supports accessibility adaptation and content localization.

Technology Category

• Search and Optimization: Evaluation and Analysis
• Machine Learning: Optimization
• Constraint Satisfaction and Optimization: Constraint Optimization

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Web Mining and Content Analysis: Web data visualization
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

Effective chart design is essential for satisfying viewers' information needs, such as retrieving values from a chart or comparing two values. However, creating effective charts is challenging and time-consuming due to the large design space and the inter-dependencies between individual design parameters. To address this challenge, we propose ChartOptimiser -- a Bayesian approach for task-driven optimisation of charts, such as bar charts. At the core of ChartOptimiser is a novel objective function to automatically optimise an eight-dimensional design space combining four perceptual metrics: visual saliency, text legibility, colour preference, and white space ratio. Through empirical evaluation on 12 bar charts and four common analytical tasks -- finding the extreme value, retrieving a value, comparing two values, and computing a derived value -- we show that ChartOptimiser outperforms existing design baselines concerning task-solving ease, visual aesthetics, and chart clarity. We also discuss two practical applications of ChartOptimiser: generating charts for accessibility and content localisation. Taken together, ChartOptimiser opens up an exciting new research direction in automated chart design where charts are optimised for users' information needs, preferences, and contexts.

Problem

• Optimizing chart designs for task efficiency and user needs
• Addressing challenges in large design space and parameter dependencies
• Improving task-solving ease, aesthetics, and clarity in charts

Innovation

• Bayesian approach for task-driven chart optimization
• Eight-dimensional design space with perceptual metrics
• Outperforms baselines in task-solving and aesthetics