Zilong Liu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Zilong Liu has a total of 60 co-authors. Here are the top co-authors by number of shared publications:
- Meilin Shi
Department of Geography and Regional Research University of Vienna Vienna Austria
7 shared publications Krzysztof Janowicz
University of California Santa Barbara California USA
6 shared publications- Gengchen Mai
SEAI Lab, Department of Geography and the Environment, University of Texas at Austin, Austin, TX, USA
5 shared publications - Ivan Majic
Department of Geography and Regional Research University of Vienna Vienna Austria
5 shared publications - Mina Karimi
University of Vienna
5 shared publications - Alexandra Fortacz
Department of Geography and Regional Research University of Vienna Vienna Austria
4 shared publications
Latest Publications
- Open Access
- 19 November 2026
- Published by Taylor & Francis in Spatial Cognition & Computation
- p. 1-40
Through the eyes of the machine: how foundation models represent vague cognitive regions
- 19 November 2026
- Published by Taylor & Francis in Spatial Cognition & Computation
- p. 1-40
Open AccessThrough the eyes of the machine: how foundation models represent vague cognitive regions
- Open Access
- 3 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 133-143
How Geographic Information Influences Vision-Language Model Judgments of Urban Scenes
- 3 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 133-143
Open AccessHow Geographic Information Influences Vision-Language Model Judgments of Urban Scenes
- Open Access
- 23 March 2026
- Published by Wiley in Transactions in GIS
- Vol. 30 (2)
Golden Gate Bridge, as Always? Eliciting Prototypical Places From Autoregressive Large Language Models via Category Production
Among the various theories of categorization, Eleanor Rosch's prototype theory stands out as both influential and contested. In contrast to the classical theory of concepts, prototype theory posits that humans conceptualize the world using category structures where exemplars vary in their degree of membership. While this theory remains well‐regarded among cognitive psychologists as an account of human cognition, we know little about whether artificial agents—particularly large language models (LLMs)—also use a similar type of categorization. Given that these models have shown to mirror human‐like cognitive abilities, an important question arises: Do they also encode categories with graded membership? This question also invites a broader ethical reflection: While substantial research has documented a lack of diversity in LLM outputs, few has realized that such bias can be understood as a graded structure in the model's internal representations of categories. In this study, we explore this possibility by examining the geographical conceptualizations of 11 autoregressive LLMs. Guided by prior work that advocated empirical, inductive testing for studying geographic categories, we employ a cognitive‐psychology experiment known as category production . It prompts a model to generate exemplars of a given category. This approach yields an extensive list of category norms (i.e., typical items associated with a category) across 297 geographic categories, revealing their graded structures both encoded within individual models and collectively. Notably, we discover a strong presence of prototypical places shared across different sampling temperatures, geographic categories, and models. Furthermore, we observe and explain a lack of geographic diversity in LLM outputs through the lens of prototype theory.
- 23 March 2026
- Published by Wiley in Transactions in GIS
- Vol. 30 (2)
Open AccessGolden Gate Bridge, as Always? Eliciting Prototypical Places From Autoregressive Large Language Models via Category Production
Among the various theories of categorization, Eleanor Rosch's prototype theory stands out as both influential and contested. In contrast to the classical theory of concepts, prototype theory posits that humans conceptualize the world using category structures where exemplars vary in their degree of membership. While this theory remains well‐regarded among cognitive psychologists as an account of human cognition, we know little about whether artificial agents—particularly large language models (LLMs)—also use a similar type of categorization. Given that these models have shown to mirror human‐like cognitive abilities, an important question arises: Do they also encode categories with graded membership? This question also invites a broader ethical reflection: While substantial research has documented a lack of diversity in LLM outputs, few has realized that such bias can be understood as a graded structure in the model's internal representations of categories. In this study, we explore this possibility by examining the geographical conceptualizations of 11 autoregressive LLMs. Guided by prior work that advocated empirical, inductive testing for studying geographic categories, we employ a cognitive‐psychology experiment known as category production . It prompts a model to generate exemplars of a given category. This approach yields an extensive list of category norms (i.e., typical items associated with a category) across 297 geographic categories, revealing their graded structures both encoded within individual models and collectively. Notably, we discover a strong presence of prototypical places shared across different sampling temperatures, geographic categories, and models. Furthermore, we observe and explain a lack of geographic diversity in LLM outputs through the lens of prototype theory.
- Top Cited Papers
- 6 February 2026
- Published by Wiley in Transactions in GIS
- Vol. 30 (1)
The KnowWhereGraph : A Large‐Scale Geo‐Knowledge Graph for Interdisciplinary Knowledge Discovery and Geo‐Enrichment
Global challenges such as food supply chain disruptions, public health crises, and natural hazard responses require access to and integration of diverse datasets, many of which are geospatial. Over the past few years, a growing number of (geo)portals have been developed to address this need. However, most existing (geo)portals are stacked by separated or sparsely connected data “silos” impeding effective data consolidation. A new way of sharing and reusing geospatial data is therefore urgently needed. In this work, we introduce KnowWhereGraph, a knowledge graph‐based data integration, enrichment, and synthesis framework that not only includes schemas and data related to human and environmental systems but also provides a suite of supporting tools for accessing this information. The KnowWhereGraph aims to address the challenge of data integration by building a large‐scale, cross‐domain, preintegrated, FAIR‐principles‐based, and AI‐ready data warehouse rooted in knowledge graphs. We highlight the design principles of KnowWhereGraph, emphasizing the roles of space, place, and time in bridging various data “silos.” Additionally, we demonstrate multiple use cases where the proposed geospatial knowledge graph and its associated tools empower decision‐makers to uncover insights that are often hidden within complex and poorly interoperable datasets.
- 6 February 2026
- Published by Wiley in Transactions in GIS
- Vol. 30 (1)
Top Cited PapersThe KnowWhereGraph : A Large‐Scale Geo‐Knowledge Graph for Interdisciplinary Knowledge Discovery and Geo‐Enrichment
Global challenges such as food supply chain disruptions, public health crises, and natural hazard responses require access to and integration of diverse datasets, many of which are geospatial. Over the past few years, a growing number of (geo)portals have been developed to address this need. However, most existing (geo)portals are stacked by separated or sparsely connected data “silos” impeding effective data consolidation. A new way of sharing and reusing geospatial data is therefore urgently needed. In this work, we introduce KnowWhereGraph, a knowledge graph‐based data integration, enrichment, and synthesis framework that not only includes schemas and data related to human and environmental systems but also provides a suite of supporting tools for accessing this information. The KnowWhereGraph aims to address the challenge of data integration by building a large‐scale, cross‐domain, preintegrated, FAIR‐principles‐based, and AI‐ready data warehouse rooted in knowledge graphs. We highlight the design principles of KnowWhereGraph, emphasizing the roles of space, place, and time in bridging various data “silos.” Additionally, we demonstrate multiple use cases where the proposed geospatial knowledge graph and its associated tools empower decision‐makers to uncover insights that are often hidden within complex and poorly interoperable datasets.
- Open Access
- 2 January 2026
- Published by Taylor & Francis in International Journal of Digital Earth
- Vol. 19 (1)
The geography of evolving concepts: quantifying spatial, temporal, and semantic consequences of ontology changes
- 2 January 2026
- Published by Taylor & Francis in International Journal of Digital Earth
- Vol. 19 (1)
Open AccessThe geography of evolving concepts: quantifying spatial, temporal, and semantic consequences of ontology changes
- Open Access
- 3 November 2025
- Published by Association for Computing Machinery (ACM)
- p. 799-803
Whose Truth? Pluralistic Geo-Alignment for (Agentic) AI
- 3 November 2025
- Published by Association for Computing Machinery (ACM)
- p. 799-803
Open AccessWhose Truth? Pluralistic Geo-Alignment for (Agentic) AI
- Open Access
- 13 May 2025
- Published by Wiley in Transactions in GIS
- Vol. 29 (3)
Operationalizing Geographic Diversity for the Evaluation of AI ‐Generated Content
The introduction and widespread use of foundation models has accelerated the necessity of identifying geographic bias in AI‐generated content. In this respect, we operationalize geographic diversity as a countermeasure. We refine the notion of geographic diversity as the quality of including data from various places and maintaining a balance across these places in both learning and generation processes. Drawing from information theory, ecology, and prior work in AI evaluation, we provide an entropy‐based definition of geographic diversity and propose to measure geographic diversity as effective numbers of places. We apply our measurement by studying generated content from six large language models, including GPT‐3.5, GPT‐4o, Mistral 7B, Mistral Large, Claude 3 Haiku, and Claude 3.5 Sonnet. Our case study reveals that prompt variations, such as modifying concept mentions or scale mentions in a user prompt, can result in more geographic diversity in their generated content. In addition, we observe that less advanced models can generate more geographically diverse content than state‐of‐the‐art ones. Furthermore, certain places dominate the generated content of these models, yet their prominence does not reflect their real‐world counterparts. Our work stresses the importance of quantifying geographic information in AI‐generated content to support GeoAI and the broader AI evaluation in the age of foundation models.
- 13 May 2025
- Published by Wiley in Transactions in GIS
- Vol. 29 (3)
Open AccessOperationalizing Geographic Diversity for the Evaluation of AI ‐Generated Content
The introduction and widespread use of foundation models has accelerated the necessity of identifying geographic bias in AI‐generated content. In this respect, we operationalize geographic diversity as a countermeasure. We refine the notion of geographic diversity as the quality of including data from various places and maintaining a balance across these places in both learning and generation processes. Drawing from information theory, ecology, and prior work in AI evaluation, we provide an entropy‐based definition of geographic diversity and propose to measure geographic diversity as effective numbers of places. We apply our measurement by studying generated content from six large language models, including GPT‐3.5, GPT‐4o, Mistral 7B, Mistral Large, Claude 3 Haiku, and Claude 3.5 Sonnet. Our case study reveals that prompt variations, such as modifying concept mentions or scale mentions in a user prompt, can result in more geographic diversity in their generated content. In addition, we observe that less advanced models can generate more geographically diverse content than state‐of‐the‐art ones. Furthermore, certain places dominate the generated content of these models, yet their prominence does not reflect their real‐world counterparts. Our work stresses the importance of quantifying geographic information in AI‐generated content to support GeoAI and the broader AI evaluation in the age of foundation models.
- Open Access
- 8 May 2025
- Published by Wiley in Transactions in GIS
- Vol. 29 (3)
Defining Concept Drift and Its Variants in Research Data Management: A Scientometric Case Study on Geographic Information Science
Just as the language used to express them, concepts constantly evolve, due to changing cultural norms, scientific progress, or major events that (re)shape our understanding of the world. For example, as humans increasingly contribute to disasters, e.g., by altering the landscape, the line between natural and anthropogenic disasters blurs. This has far‐reaching consequences for disaster response, policymaking, and also research. In research data management (RDM), evolving terms like natural disaster or global warming bring challenges for data retrieval, integration, interpretation, and reuse. This highlights the importance of studying concept drift, defined as the changes in a concept's meaning over time. While concept drift has been explored in numerous domains, including cognitive science, linguistics, semiotics, and ontology evolution, a clear definition within RDM is still lacking. This is significant as it may reduce recall during RDM‐based data retrieval. To bridge this gap, this work proposes definitions of concept drift and related phenomena, such as shifts and merges, to foster cross‐domain consensus. Using a scientometric dataset in GIScience, we apply these definitions to identify representative cases of concept drift. Our findings reveal that concepts in the GIScience domain may undergo various changes, including birth, drift, and shift, among others. These findings emphasize the need to monitor concept drift in RDM to uphold FAIR data principles, thereby improving precision and recall when searching for relevant literature or data. Lastly, we highlight the potential to measure concept drift over space (place) in addition to time alone.
- 8 May 2025
- Published by Wiley in Transactions in GIS
- Vol. 29 (3)
Open AccessDefining Concept Drift and Its Variants in Research Data Management: A Scientometric Case Study on Geographic Information Science
Just as the language used to express them, concepts constantly evolve, due to changing cultural norms, scientific progress, or major events that (re)shape our understanding of the world. For example, as humans increasingly contribute to disasters, e.g., by altering the landscape, the line between natural and anthropogenic disasters blurs. This has far‐reaching consequences for disaster response, policymaking, and also research. In research data management (RDM), evolving terms like natural disaster or global warming bring challenges for data retrieval, integration, interpretation, and reuse. This highlights the importance of studying concept drift, defined as the changes in a concept's meaning over time. While concept drift has been explored in numerous domains, including cognitive science, linguistics, semiotics, and ontology evolution, a clear definition within RDM is still lacking. This is significant as it may reduce recall during RDM‐based data retrieval. To bridge this gap, this work proposes definitions of concept drift and related phenomena, such as shifts and merges, to foster cross‐domain consensus. Using a scientometric dataset in GIScience, we apply these definitions to identify representative cases of concept drift. Our findings reveal that concepts in the GIScience domain may undergo various changes, including birth, drift, and shift, among others. These findings emphasize the need to monitor concept drift in RDM to uphold FAIR data principles, thereby improving precision and recall when searching for relevant literature or data. Lastly, we highlight the potential to measure concept drift over space (place) in addition to time alone.
- Top Cited PapersOpen Access
- 3 April 2025
- Published by Taylor & Francis in Cartography and Geographic Information Science
- Vol. 52 (4) , 331-349
Geography for AI sustainability and sustainability for GeoAI
- 3 April 2025
- Published by Taylor & Francis in Cartography and Geographic Information Science
- Vol. 52 (4) , 331-349
Top Cited PapersOpen AccessGeography for AI sustainability and sustainability for GeoAI
- Open Access
- 11 December 2024
- Published by Elsevier in Journal of Web Semantics
- Vol. 84
The KnowWhereGraph ontology
- 11 December 2024
- Published by Elsevier in Journal of Web Semantics
- Vol. 84
Open AccessThe KnowWhereGraph ontology