Exploring User Evaluations of Visualization of Chart Features of Experimental Data

Data visualization is essential for turning data into meaningful insights. This is highlighted in the context of experimental data for scientific research purpose due to the complexity and scale of the data involved. In this paper, we investigated the relationship between the visualization effect of experimental data and chart features. We analyzed the effects of chart axes, lines, text display ratio and content display ratio on the visualization. The chart parameter model and chart content model were proposed to capture the effect of different features on the degree of visualization and to explore the cross-influence of different models on the degree of visualization. This study also introduced machine learning models to help build a set of evaluation models for a more readable visualization for experimental data. This work may help authors to set the chart parameters into the paper more easily, as well as improve the efficiency of the readers’ reading comprehension of scientific figures.