Abstract: Information overload refers to the difficulty in understanding information due to an inappropriate density of information in a limited display space. Scatter plots, being one of the most common data visualization methods, are particularly susceptible to information overload and require techniques to mitigate this issue. From the perspective of visual analysis, we put forward a clutter entropy model for quantitative evaluation of scatter plots information clutter, and propose an algorithm to eliminate scatter plots information clutter from the two dimensions of content and space based on individual significance model and quad tree structure. The high and low significance data are processed and the display size and layout position are specified respectively, and the scatter plots information clutter is eliminated from the display and layout level.