580 California St., Suite 400
San Francisco, CA, 94104
Key research themes
1. How can multi-criteria spatial preference queries be efficiently modeled and solved in spatial data systems?
This research area explores computational methods to rank and select spatial objects based on user-defined quality preferences within their spatial neighborhoods. The focus is on defining flexible spatial preference models, integrating heterogeneous attribute sources, and developing efficient algorithms to process top-k queries over spatial data considering proximity-weighted and functional neighborhood definitions. Such methods are crucial for applications like real estate selection, urban planning, and location-based decision-making where location quality assessment depends on complex multi-attribute spatial interactions.
Key finding: This paper formulates spatial preference queries as top-k ranking problems where spatial objects are scored based on the quality of features in their spatial neighborhoods defined through either explicit circular regions or... Read more
2. What are the best strategies and algorithms for solving spatial facility location and layout optimization problems considering real-world constraints?
This theme covers the mathematical modeling, algorithm design, and practical applications of facility location and layout problems in spatial domains under varying constraints such as user demand dispersion, resource capacities, spatial contiguity, and multi-level environments. It emphasizes approaches that ensure minimal delivery distances, robustness, and efficiency while also considering human-centric parameters, sustainability, and integration with crowd-sourced or home-based services. Advances in heuristic and metaheuristic algorithms, as well as p-median problem formulations, are central to improving logistics and supply chain performance.
Key finding: This work develops a hybrid heuristic-metaheuristic model combining a greedy algorithm and simulated annealing to optimally locate user household collection and delivery points (CDPs) by solving a p-median problem. The model... Read more
Key finding: This paper introduces 'spopt,' an open-source Python package implemented within PySAL for solving diverse spatial optimization problems including regionalization and facility location models such as Max-p regions and p-median... Read more
Key finding: The paper presents the Origin (TO) algorithm, a bottom-left-filled heuristic developed to efficiently optimize manufacturing plant facility layouts considering sustainability and material handling costs. By assuming... Read more
Key finding: This conceptual work bridges spatial statistics—particularly spatial autocorrelation and spatial median computations—with spatial optimization problems like the p-median location problem. It highlights recursive algorithms... Read more
Key finding: The authors develop a Bayesian spatial interaction model (BSIM) that probabilistically estimates location-specific revenues for competitive facility location planning. Integrating factors like facility attractiveness and... Read more
3. How can spatial attention, working memory, and neural representations of space be connected to understand spatial prioritization and navigation?
This research investigates cognitive and neural mechanisms of spatial attention, spatial working memory, and spatial navigation, focusing on how attentional selection operates both perceptually and internally, and how spatial representations in the brain support navigation in complex environments. It includes experimental dissociations of perceptual and memory-based spatial prioritization, neural correlates bridging parietal cortical and hippocampal activity, and theoretical taxonomies linking behavior to neural coding. Understanding these mechanisms is essential for elucidating spatial prioritization at behavioral and neuronal levels.
Key finding: Through spatial working memory updating tasks with manipulated perceptual attention, the study shows that attentional selection in working memory operates independently from perceptual selective attention mechanisms,... Read more
Key finding: Neuroscientific recordings from macaque posterior parietal cortex (PPC) and hippocampus during virtual maze navigation reveal substantial position-selective neural activity in PPC, not solely explained by saccades or visual... Read more
Key finding: The authors propose a comprehensive taxonomy of mammalian spatial navigation processes, synthesizing behavioral and neural data. They highlight multiple navigation strategies (path following, landmark usage, triangulation,... Read more
4. How can spatial preference and risk analyses incorporate structured preference functions and multi-attribute utility to support spatial decision making under uncertainty?
Research in this area develops mathematical theories and preference models to represent decision-maker values over spatially distributed outcomes, including risk and uncertainty. It focuses on measurable spatial value and utility functions formalized via preference conditions such as spatial preferential independence and homogeneity. These models enable spatial decision support through additive and utility-based functions that capture spatial attributes across regions, improving decision-making frameworks in environmental management, resource allocation, and risk analyses.
Key finding: This theoretical paper establishes preference conditions that guarantee the existence of additive measurable spatial value functions and utility functions to represent ordinal preferences over multi-attribute spatial... Read more