A learning-based algorithm for generation of synthetic participatory mapping data in 2D and 3D.

Area of Science:

• Geographic Information Systems (GIS)
• Spatial Analysis
• Environmental Science

Background:

• Public Participation GIS (PPGIS) collects experiential spatial knowledge through map-based surveys.
• PPGIS data, often point-based, are crucial for urban and environmental research.
• Collecting PPGIS data solely for methodological development can be resource-intensive.

Purpose of the Study:

• To develop a novel context-aware method for generating synthetic PPGIS data.
• To create realistic 2D and 3D PPGIS datasets for methodological advancement.
• To support the development of spatial analytical techniques for emerging 3D PPGIS.

Main Methods:

• A context-aware machine learning approach was employed.
• The method learns from existing 2D and 3D PPGIS data and geographical context.
• Simulated PPGIS point datasets are generated for specified geographical extents.

Main Results:

• A realistic, context-aware synthetic PPGIS point dataset can be generated.
• The method supports the creation of both 2D and 3D spatial data.
• Facilitates methodological development in the absence of extensive real-world data.

Conclusions:

• The developed method provides a cost-effective way to obtain PPGIS data for research.
• Enables advancements in spatial analysis, particularly for 3D PPGIS.
• Highlights the potential of synthetic data for supporting emerging GIS technologies.