Authors

Yuanzhe Jin

Abstract

When visualizing large-scale or high-density datasets, overplotting in scatter plots can severely hinder data exploration, as overlapping points obscure underlying patterns and distributions, affecting users’ ability to interpret the content. To address this issue, we propose HistoScatter, a novel hybrid visualization technique that integrates scatter plots and histograms to tackle this challenge. We evaluate HistoScatter through experiments and user evaluations on real-world datasets. The results show that, compared to traditional scatter plots and some existing solutions, HistoScatter achieves superior user satisfaction and visual effectiveness. HistoScatter effectively balances the trade-off between information density and visual clarity, offering a practical solution for large-scale data visualization in fields such as data science, business intelligence, and exploratory data analysis. Download paper here