I-GUIDE Platform for Geospatial Data-Intensive Knowledge Discovery through Scalable Knowledge Graph and Object Storage - BYOP
Fuente:
Zenodo
Gespeichert in:
| Hauptverfasser: | , , , , , , |
|---|---|
| Format: | Recurso digital |
| Sprache: | Englisch |
| Veröffentlicht: |
Zenodo
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866901371800780800 |
|---|---|
| author | Erick, Li Kumar, Arunesh Baig, Furqan Kang, Yunfan Jaroenchai, Nattapon Padmanabhan, Anand Wang, Shaowen |
| author_facet | Erick, Li Kumar, Arunesh Baig, Furqan Kang, Yunfan Jaroenchai, Nattapon Padmanabhan, Anand Wang, Shaowen |
| contents | <p>The I-GUIDE Platform provides a scalable, user-centric online environment designed to support geospatial data-intensive convergence research and education. This science gateway enables knowledge discovery, cross-disciplinary collaboration, and computational reproducibility through integrated data infrastructure and AI tools.<br><br>The platform consists of three primary components: a React-based frontend, a Node.js backend, and a Neo4j graph database. All components are hosted on the Jetstream2 cloud infrastructure that is supported by the National Science Foundation. The platform supports dedicated data hosting capabilities through MinIO–a high-performance, S3-compatible object storage system–ensuring reliable storage of diverse research files, including large unstructured datasets and CyberGIS-Jupyter notebooks. Importantly, DOI can be issued for user-uploaded datasets, facilitating long-term discoverability and citation of data-intensive scientific knowledge.<br><br>Complementing spatially-aware discovery, OpenSearch is used for indexing and querying structured metadata and GeoJSON-based spatial descriptors. This enables the Element Map feature, which visualizes elements on an interactive map displaying their locations and boundaries. This allows users to intuitively explore spatial relationships between knowledge elements.<br><br>The platform’s user interface provides seamless navigation, contribution, and collaboration capabilities, enabling users to access and manage spatially indexed content with ease. To further enhance user experience, the platform introduces a smart search chatbot powered by a Retrieval-Augmented Generation (RAG) pipeline. The chatbot leverages Large Language Models (LLMs) and spatial metadata to provide context-aware, intelligent responses grounded in geospatial knowledge.<br><br>The platform also integrates a JupyterHub environment, enabling users to run uploaded notebooks using high-performance and cloud computing resources with pre-installed geospatial packages. Upcoming support for GPU-enabled HPC resources will further enhance performance for large-scale data analysis and modeling.<br><br>Platform stability is maintained through an automated testing pipeline, while an administrative dashboard facilitates content curation and user activity oversight, ensuring high-quality knowledge contributions and data integrity.<br><br>Users engage with the platform through a variety of use cases, including accessing categorized knowledge elements such as maps, datasets, Jupyter notebooks, publications, educational resources, and GitHub code repositories. The LLM-enabled smart search facilitates exploratory workflows by allowing natural language queries, thereby supporting intuitive access to relevant resources and results. To promote discovery and contextual understanding, users can navigate related elements through the Element Map. Researchers can execute Jupyter notebooks directly within the platform via integrated JupyterHub support, enabling reproducible analysis. Additionally, the platform supports community-driven knowledge exchange, allowing users to contribute original or curated resources through submission forms and data upload interfaces, with the added option to generate DOIs for publishing datasets through the platform.<br><br>Together, these user-centric capabilities and core functionalities form a cohesive, extensible science gateway that empowers researchers and educators to share, explore, and analyze geospatial data for various convergence research and education purposes. By combining cloud-based storage, spatial metadata, AI-powered search, and computational tools at scale, the I-GUIDE Platform advances geospatial data-intensive convergence research, education, and innovation.<br><br>Portal URL: https://platform.i-guide.io</p> |
| format | Recurso digital |
| id | zenodo_https___doi_org_10_5281_zenodo_17450574 |
| institution | Zenodo |
| language | eng |
| publishDate | 2025 |
| publisher | Zenodo |
| record_format | zenodo |
| spellingShingle | I-GUIDE Platform for Geospatial Data-Intensive Knowledge Discovery through Scalable Knowledge Graph and Object Storage - BYOP Erick, Li Kumar, Arunesh Baig, Furqan Kang, Yunfan Jaroenchai, Nattapon Padmanabhan, Anand Wang, Shaowen cyberGIS knowledge graph convergence science geospatial ai science gateway <p>The I-GUIDE Platform provides a scalable, user-centric online environment designed to support geospatial data-intensive convergence research and education. This science gateway enables knowledge discovery, cross-disciplinary collaboration, and computational reproducibility through integrated data infrastructure and AI tools.<br><br>The platform consists of three primary components: a React-based frontend, a Node.js backend, and a Neo4j graph database. All components are hosted on the Jetstream2 cloud infrastructure that is supported by the National Science Foundation. The platform supports dedicated data hosting capabilities through MinIO–a high-performance, S3-compatible object storage system–ensuring reliable storage of diverse research files, including large unstructured datasets and CyberGIS-Jupyter notebooks. Importantly, DOI can be issued for user-uploaded datasets, facilitating long-term discoverability and citation of data-intensive scientific knowledge.<br><br>Complementing spatially-aware discovery, OpenSearch is used for indexing and querying structured metadata and GeoJSON-based spatial descriptors. This enables the Element Map feature, which visualizes elements on an interactive map displaying their locations and boundaries. This allows users to intuitively explore spatial relationships between knowledge elements.<br><br>The platform’s user interface provides seamless navigation, contribution, and collaboration capabilities, enabling users to access and manage spatially indexed content with ease. To further enhance user experience, the platform introduces a smart search chatbot powered by a Retrieval-Augmented Generation (RAG) pipeline. The chatbot leverages Large Language Models (LLMs) and spatial metadata to provide context-aware, intelligent responses grounded in geospatial knowledge.<br><br>The platform also integrates a JupyterHub environment, enabling users to run uploaded notebooks using high-performance and cloud computing resources with pre-installed geospatial packages. Upcoming support for GPU-enabled HPC resources will further enhance performance for large-scale data analysis and modeling.<br><br>Platform stability is maintained through an automated testing pipeline, while an administrative dashboard facilitates content curation and user activity oversight, ensuring high-quality knowledge contributions and data integrity.<br><br>Users engage with the platform through a variety of use cases, including accessing categorized knowledge elements such as maps, datasets, Jupyter notebooks, publications, educational resources, and GitHub code repositories. The LLM-enabled smart search facilitates exploratory workflows by allowing natural language queries, thereby supporting intuitive access to relevant resources and results. To promote discovery and contextual understanding, users can navigate related elements through the Element Map. Researchers can execute Jupyter notebooks directly within the platform via integrated JupyterHub support, enabling reproducible analysis. Additionally, the platform supports community-driven knowledge exchange, allowing users to contribute original or curated resources through submission forms and data upload interfaces, with the added option to generate DOIs for publishing datasets through the platform.<br><br>Together, these user-centric capabilities and core functionalities form a cohesive, extensible science gateway that empowers researchers and educators to share, explore, and analyze geospatial data for various convergence research and education purposes. By combining cloud-based storage, spatial metadata, AI-powered search, and computational tools at scale, the I-GUIDE Platform advances geospatial data-intensive convergence research, education, and innovation.<br><br>Portal URL: https://platform.i-guide.io</p> |
| title | I-GUIDE Platform for Geospatial Data-Intensive Knowledge Discovery through Scalable Knowledge Graph and Object Storage - BYOP |
| topic | cyberGIS knowledge graph convergence science geospatial ai science gateway |
| url | https://doi.org/10.5281/zenodo.17450574 |