Data Visualization

Data visualization aims to translate complex datasets into readily understandable visual representations, facilitating insights and informed decision-making. Current research emphasizes automating visualization generation from natural language queries, leveraging large language models (LLMs) and incorporating user preferences into multidimensional projection techniques like t-SNE and UMAP, often enhanced by manifold learning methods. These advancements improve accessibility for non-experts and enhance the interpretability of visualizations across diverse fields, from oceanography and finance to machine learning explainability. The development of comprehensive benchmarks and interactive tools further strengthens the field's capacity for rigorous evaluation and user-centered design.