On November 4th, a team led by Professor Li Zhilin from the Faculty of Geoscience and Engineering published their academic paper, “Point-Feature Label Placement with Maximum Entropy Principle,” in International Journal of Geographical Information Science. Wu Zhiwei, a PhD student of 2025, was the first author, with Associate Professor Lan Tian serving as the corresponding author.

Title page of the paper

According to the research, the user experience of a map can be greatly enhanced through high-quality label placement. This study introduces a new method for annotating point elements, combining the maximum entropy principle with an improved genetic algorithm. Comparative experiments have been conducted with 29 existing labelling methods in a dataset of 1000 points, and the results indicate that the average percentage of non-overlapping labels by the proposed method reaches 94.73% (i.e., being ranked second in all 30 methods), between which no statistically significant difference is noted from the best one (i.e., 94.87%). Further analysis shows that the proposed method excels in prioritizing label positions and could be extended to place labels for line and area features. This research offers fresh approaches to optimizing map label placement.

Link: https://www.tandfonline.com/eprint/J86GYVCQQ4ZGQYHIGP5S/full?target=10.1080/13658816.2024.2422502