Advancing Large Language Models for Spatiotemporal and Semantic Association Mining of Similar Environmental Events DOI Open Access
Yuanyuan Tian, Wenwen Li, Lei Hu

et al.

Transactions in GIS, Journal Year: 2024, Volume and Issue: unknown

Published: Dec. 5, 2024

ABSTRACT Retrieval and recommendation are two essential tasks in modern search tools. This paper introduces a novel retrieval‐reranking framework leveraging large language models to enhance the spatiotemporal semantic associated mining of relevant, unusual climate environmental events described news articles web posts. uses advanced natural processing techniques address limitations traditional manual curation methods terms high labor costs lack scalability. Specifically, we explore an optimized solution employ cutting‐edge embedding for semantically analyzing (news) propose Geo‐Time Re‐ranking strategy that integrates multi‐faceted criteria including spatial proximity, temporal association, similarity, category‐instructed similarity rank identify similar events. We apply proposed dataset four thousand local observer network events, achieving top performance on recommending among multiple dense retrieval models. The pipeline can be applied wide range data dealing with geospatial data. hope by linking relevant better aid general public gain enhanced understanding change its impact different communities.

Language: Английский

Geospatial foundation models for image analysis: evaluating and enhancing NASA-IBM Prithvi’s domain adaptability DOI

Chia-Yu Hsu,

Wenwen Li, Sizhe Wang

et al.

International Journal of Geographical Information Science, Journal Year: 2024, Volume and Issue: unknown, P. 1 - 30

Published: Aug. 30, 2024

Research on geospatial foundation models (GFMs) has become a trending topic in artificial intelligence (AI) research due to their potential for achieving high generalizability and domain adaptability, reducing model training costs individual researchers. Unlike large language models, such as ChatGPT, constructing visual image analysis, particularly remote sensing, encountered significant challenges formulating diverse vision tasks into general problem framework. This paper evaluates the recently released NASA-IBM GFM Prithvi its predictive performance high-level analysis across multiple benchmark datasets. was selected because it is one of first open-source GFMs trained time-series high-resolution sensing imagery. A series experiments were designed assess Prithvi's compared other pre-trained task-specific AI analysis. New strategies, including band adaptation, multi-scale feature generation, fine-tuning techniques, are introduced integrated an pipeline enhance adaptation capability improve performance. In-depth analyses reveal strengths weaknesses, offering insights both improving developing future tasks.

Language: Английский

Citations

4
Open science 2.0: revolutionizing spatiotemporal data sharing and collaboration DOI Creative Commons
Siqin Wang, Xiaoxiao Huang, Mengxi Zhang

et al.

Computational Urban Science, Journal Year: 2025, Volume and Issue: 5(1)

Published: Jan. 26, 2025

Abstract The Spatial Data Lab (SDL) project is a collaborative initiative by the Center for Geographic Analysis at Harvard University, KNIME, Future Lab, China Institute, and George Mason University. Co-sponsored NSF IUCRC Spatiotemporal Innovation Center, SDL aims to advance applied research in spatiotemporal studies across various domains such as business, environment, health, mobility, more. focuses on developing an open-source infrastructure data linkage, analysis, collaboration. Key objectives include building services, reproducible, replicable, expandable (RRE) platform, workflow-driven analysis tools support case studies. Additionally, promotes science training, cross-party collaboration, creation of geospatial that foster inclusivity, transparency, ethical practices. Guided academic advisory committee world-renowned scholars, laying foundation more open, effective, robust scientific enterprise.

Language: Английский

Citations

0
Advancing Arctic Sea Ice Remote Sensing with AI and Deep Learning: Opportunities and Challenges DOI Creative Commons
Wenwen Li, Chia-Yu Hsu, Marco Tedesco

et al.

Remote Sensing, Journal Year: 2024, Volume and Issue: 16(20), P. 3764 - 3764

Published: Oct. 10, 2024

Revolutionary advances in artificial intelligence (AI) the past decade have brought transformative innovation across science and engineering disciplines. In field of Arctic science, we witnessed an increasing trend adoption AI, especially deep learning, to support analysis big data facilitate new discoveries. this paper, provide a comprehensive review applications learning sea ice remote sensing domains, focusing on problems such as lead detection, thickness estimation, concentration extent forecasting, motion type classification. addition discussing these applications, also summarize technological that customized solutions, including loss functions strategies better understand dynamics. To promote growth exciting interdisciplinary field, further explore several research areas where community can benefit from cutting-edge AI technology. These include improving multimodal capabilities, enhancing model accuracy measuring prediction uncertainty, leveraging foundation models, deepening integration with physics-based models. We hope paper serve cornerstone progress using inspire field.

Language: Английский

Citations

3
Advancing Large Language Models for Spatiotemporal and Semantic Association Mining of Similar Environmental Events DOI Open Access
Yuanyuan Tian, Wenwen Li, Lei Hu

et al.

Transactions in GIS, Journal Year: 2024, Volume and Issue: unknown

Published: Dec. 5, 2024

ABSTRACT Retrieval and recommendation are two essential tasks in modern search tools. This paper introduces a novel retrieval‐reranking framework leveraging large language models to enhance the spatiotemporal semantic associated mining of relevant, unusual climate environmental events described news articles web posts. uses advanced natural processing techniques address limitations traditional manual curation methods terms high labor costs lack scalability. Specifically, we explore an optimized solution employ cutting‐edge embedding for semantically analyzing (news) propose Geo‐Time Re‐ranking strategy that integrates multi‐faceted criteria including spatial proximity, temporal association, similarity, category‐instructed similarity rank identify similar events. We apply proposed dataset four thousand local observer network events, achieving top performance on recommending among multiple dense retrieval models. The pipeline can be applied wide range data dealing with geospatial data. hope by linking relevant better aid general public gain enhanced understanding change its impact different communities.

Language: Английский

Citations

0