UC Riverside

Spatial data analysis plays a crucial role in understanding the complex dynamics of our world across various fields including geography, urban planning, environmental science, and transportation. As spatial data becomes more intricate due to the advances in geospatial technologies, there is an urgent need for more efficient and expressive tools and techniques for its analysis and modeling. This dissertation addresses this demand, presenting a series of interconnected research contributions that have enhanced the landscape of spatial data analysis.

Firstly, the dissertation introduces the opensource Python Library Pyneapple. Pyneapple encapsulates the aforementioned tools and techniques, providing a platform for scalable and expressive spatial data analysis. Secondly, among the components of the Pynapple, we highlight the algorithms for solving the "Expressive Max-p regions" (EMP) problem and the Spatial Network Hotspot Detection (SNHD). EMP is a key component of the regionalization module of Pyneapple. It expands the traditional max-p regions problem by supporting multiple constraints with range operators, which allows for more precise and enriched spatial analysis. SNHD illustrates how to identify statistically robust spatial hotspots efficiently and effectively, contributing to various practical fields, including transportation analysis and crime detection. Subsequently, it introduces "CleanUpOurworld," a comprehensive and interactive database for visualizing global litter data. This tool illuminates the global litter problem by making spatial data accessible and easy to understand at various spatial scales. Following this, the dissertation presents "HiDAM," a novel, research-friendly data model for high-definition maps that offers detailed and nuanced analysis of both on-road and off-road spatial data.

Collectively, these works embody a significant stride in spatial data analysis, offering a more nuanced understanding of complex spatial data and its practical applications. By developing more advanced tools and extending the existing techniques, this dissertation presents meaningful and impactful contributions to the field of spatial data analysis.