Integrating Scientific GUI Applications into Autoscaled Science Gateway Environments

The Open Metadata Exchange: A decentralized network supporting FAIR (meta)data practices and persistence for science gateways Anoop Aryal ISKME Half Moon Bay, CA, USA [email protected]g Christian Clauss ISKME Half Moon Bay, CA, USA [email protected] Sam Donovan BioQUEST Curriculum Consortium Raymond, NH, USA https://orcid.org/0000-0003-4021-9878 M. Drew LaMar Department of Biology William & Mary Williamsburg, VA, USA https://orcid.org/0000-0002-4037-1848 Anjali Nambiar ISKME Half Moon Bay, CA, USA [email protected]g Lisa Petrides ISKME Half Moon Bay, CA, USA [email protected]g Abstract—The integration of FAIR (Findability, Accessibility, Interoperability, and Reusability) standards in both open science and open education holds transformative potential for advancing STEM education and broadening participation in scientific research. Despite the well-documented benefits of the open practices associated with FAIR standards, adoption remains limited. Science gateways are a particularly attractive area for FAIR implementation, as they form a distributed network of userfriendly web portals supporting access for domain researchers to high-performance computing, scientific data, software, and open educational resources. This paper introduces the Open Metadata Exchange (OME), a decentralized network of repositories enabling the sharing of content metadata and resources. The OME network provides a novel solution to FAIRification challenges associated with the distributed nature of science gateways by directly addressing resource persistence in the context of sustainability and decommissioning. Index Terms—OER, metadata, FAIR practices, decentralized networks I. INTRODUCTION The demand for, principles guiding, and productivity promises of open science have been well documented [1]. Still, awareness and adoption of FAIR standards (Findability, Accessibility, Interoperability, and Reusability) within research communities has been limited at best [2]–[4]. Recognizing common features across open science and open education, the United Nations Educational, Scientific, and Cultural Organization (UNESCO) has recommended that standards for open educational resources (OER) be embedded into policy frameworks that cut across open access, open data, and open source software [5]. UNESCO defines OER as learning, teaching, and research materials in any format and medium that reside in the public domain or are under copyright that have been released National Science Foundation Award #2418329. under an open license, which permit no-cost access, re-use, repurposing, adaptation, and redistribution. Open environments require the use of FAIR standards to operate at scale and maximize use by target communities [6]. Science gateways are community-developed cyberinfrastructure that use FAIR standards to describe open data, software, computational infrastructure, and open educational resources using federated relational databases, and interoperate across platforms connecting users with open materials in the context of their professional work [7]–[9]. Numerous other organizations, including federally funded repositories, education service providers, and university libraries, are also involved in the creation, curation, and maintenance of open science and education materials. Robust integration across heterogeneous repositories in open science and education requires the development of shared standards for describing, finding, accessing, and distributing open materials. Working across open education and science gateway platforms involves additional challenges to the implementation of FAIR and open principles in undergraduate STEM education. The isolated nature of these standalone repositories can lead to librarians, gateway curators, educators, and students having to navigate multiple libraries and resources while duplicating the efforts to discover, tag, curate, and expand their collections. Additional challenges, with direct references to FAIR practices, include the following: •Many repositories, especially ones managed by science gateways, are not indexed in a “global” searchable database, making it difficult to find relevant material for their users outside of their gateway. Additionally, the isolated nature of many repositories makes discoverability and dissemination difficult (Findability). •Sustainability of science gateways is a long-term challenge [10], [11], with many open science and education materials and metadata being lost when a site needs to be decommissioned due to loss of funding or management (Accessibility/Preservability). •Repository metadata might have alignment issues due to differing standards, making it difficult for managers to integrate and curate material across repositories (Interoperability). •The immense volume of material across platforms makes it difficult to assess credibility, quality and relevance of content material for reusability in different contexts (Reusability). •Existing open search and indexing systems become a single point of failure as a central aggregator of metadata across open repositories. II. THE OPEN METADATA EXCHANGE As a proposed solution to these challenges, the Open Metadata Exchange (OME) [12] was created as a decentralized network of participating repositories and libraries, enabling the sharing of content (meta)data in a cooperative network for search and discovery (see Fig. 1). Any repository (Fig. 1, grey boxes) can join an existing “node” (Fig. 1, blue boxes) or host its own node in the OME network, contributing its resources – subject to any customizable restrictions or permissions – while also gaining access to materials shared by other nodes. The OME is developed and maintained independently of the content management system (CMS) where the repository might be managed, facilitating broad compatibility and uptake. For each repository or CMS, bidirectional communication with the OME is achieved through the development of an API plugin (Fig. 1, dashed arrows) that publishes resources to the OME and ingests resources from the OME back into the repository. The library of plugins will themselves be curated over time and freely open and available for use, reducing further the barrier-to-entry for repositories to join the OME Network. The current library of plugins includes support for ERIC (https://eric.ed.gov), OER Commons (https://oercommons.org), OpenLibrary (https://openlibrary.org), QUBES (https://qubeshub.org) (and more broadly HubZero CMS (https://hubzero.org)), the World Historical Gazetteer (https://whgazetteer.org), and the Internet Archive (https://archive.org). Each node in the OME consists of two parts: a peer-to-peer (P2P) (meta)data transport component (Fig. 1, solid arrows) running an InterNetNews server [13], [14] (a Usenet news server utilizing the Network News Transport Protocol (NNTP) [15]) and a front-end React UI component for searching, browsing, and curating materials within the OME Network. Each node will have the freedom to customize the level of resource access. Once a repository manager finds a resource in the OME that they would like to incorporate, they can choose to ingest just the metadata (thereby not duplicating the resource itself), or the entire resource (if made available). More specifically, the OME nodes consist of Docker containers with three layers (see Fig. 2): a front-end (FE) server powered Fig. 1. The OME Network consists of customizable “nodes” (blue boxes) with two components: a P2P data transport component running an InterNetNews server [13], [14] (solid arrows) and the OME Software. The OME Software consists of a front-end React UI for search and browse of the OME, and an API plugin (dashed arrows) to connect to servers/repositories (grey boxes) for (meta)data importing/exporting. by Node.js and a React app, and a FastAPI [16] middle layer connecting the FE server with the P2P InterNetNews backend that stores the metadata and transfers them between nodes. A. Current Status and Future Plans The OME is currently in active development. It was created by members of the Institute for the Study of Knowledge Management in Education (ISKME; https://www.iskme.org/), in partnership with and funding from the Institute for Museum and Library Services (IMLS; https://www.imls.gov/). ISKME is the creator of OER Commons, with a mission to make learning and knowledge sharing more participatory and open for all, and works to instill principles of collaborative instructional practice and curriculum enhancement. Their background in supporting over two dozen libraries of open educational resources, including the National Science Digital Library (NSDL; https://oercommons.org/hubs/NSDL) when it was defunded by NSF, ideally positions ISKME to oversee the collaborative development of the decentralized OME framework. The OME project has completed the two initial development stages of (1) development of the open-source software component creating the OME, and (2) development of a prototype OME Network for demonstration purposes. There is currently Fig. 2. Detailed diagram of OME Docker containerized node architecture. The initial layer is an FE server powered by Node.js and a React app. The middle layer is the FastAPI server, connecting the P2P InterNetNews backend, which houses and transports the metadata across nodes, to the FE server. active outreach and engagement with potential partners interested in collaborating, either as software developers, governing partners, or integrating themselves as nodes to expand the OME Network. An important future step is the establishment of a data governance structure for long-term sustainability. The goal is for ISKME to lead this effort until an independent, elected governing body is established, at which point ISKME will become a stakeholder. Examples of concerns the governing body will address include (1) establishing clear terms and conditions for network participation, with enforcement mechanisms for violation, (2) holding regular policy reviews to ensure compliance and relevance, and (3) leading communitydriven quality control mechanisms. III. DISCUSSION The OME shares many of the typical advantages of decentralized networks, in particular resilience (each node operates independently of the other nodes, safeguarding against data loss), scalability (resource allocation based on need), financial equity (cost-sharing), and flexibility (nodes can innovate and customize features as long as peering protocol is stable). Besides these advantages of decentralization, there are additional benefits that the OME offers in comparison to typical API connections between servers (see Table I). While initially envisioned as a way to share open educational resources across the landscape of open repositories, it became apparent that the OME could be used for scientific (meta)data more broadly. We believe that this can be of particular interest to the science gateways community, with its large network of web portals spanning multiple scientific disciplines, and housing materials consisting of workflows utilizing highperformance computing, scientific data, software, and open educational resources for educational outreach. Additionally, the OME has the potential to address the long-standing issue TABLE I BENEFITS OF OME COMPARED WITH TYPICAL API CONNECTIONS (HUB-AND-SPOKE MODEL). Category OME Typical API connections Development Effort Plugin for OME to/from platform. Exchange data with all participants in the network. Each pair of participants need to develop integration with each other. May not be bidirectional. Operational redundancy Completely decentralized. No single entity is vital to continued operations. Individual entities are vital for integrations to function. Operational availability/dependency All functionality is contained within the local OME node. Availability is not dependent on any other node other than your own. Availability is dependent on the API provider. Data redundancy Nodes utilize storeand-forward method to propagate data. Each node can have complete copy of everything on the network. If a participating node ceases operations, data is not lost. If the API provider ceases operations, further access to data requires manual intervention. Scalability Since all functionality is contained within your own node, you can scale the node to suit your own needs. Not dependent on the resources made available by others. Each entity has to plan and allocate enough resources to meet the needs of others utilizing their API. of sustainability and data loss when a gateway has to shut down due to lack of funding and support. The OME will be particularly useful for gateway curators, especially those working at the interface of multiple scientific disciplines where it can be difficult to create, manage, and collect useful resources across seemingly disparate areas. For example, it has the potential to strengthen the connections between education and science service providers through OER that utilizes modern computational and data-centric tools [17]. The sharing of metadata across gateways will also help with dissemination, as curators can integrate a resource into their library without duplication, making sure that the originating resource gets the proper attribution and web visits. Additionally, each curator can incorporate only the relevant metadata for their gateway, adding additional details as needed for their web portal’s unique context. Finally, and probably most importantly, this will provide enormous benefit to scientists, educators, curriculum designers, and students, as the OME increases findability of resources for research, teaching, and learning. ACKNOWLEDGMENTS This material is based upon work supported by the National Science Foundation under Grant No. 2418329. QUBES would like to acknowledge NSF IUSE 1446258, 1446269, and 1446284, and the Hewlett Foundation. ISKME would like to acknowledge the Institute for Museum and Library Services for pilot funding. REFERENCES [1] National Science Foundation, “NSF’S PUBLIC ACCESS PLAN: Today’s Data, Tomorrow’s Discoveries Increasing Access to the Results of Research Funded by the National Science Foundation,” 2015, https://www.nsf.gov/pubs/2015/nsf15052/nsf15052.pdf. [2] J. Brock, “A love letter to your future self: what scientists need to know about FAIR data,” Nature Index, 2019, https://www.natureindex.com/news-blog/what-scientists-need-to-knowabout-fair-data. [3] B. Mons, E. Schultes, F. Liu, and A. Jacobsen, “The FAIR principles: First generation implementation choices and challenges,” Data Intelligence, 2(1-2), 2020, pp.1–9. [4] M.D. Wilkinson, M. Dumontier, I.J. Aalbersberg, G. Appleton, M. Axton, A. Baak, N. Blomberg, J.W. Boiten, L.B. da Silva Santos, P.E. Bourne, and J. Bouwman, “The FAIR Guiding Principles for scientific data management and stewardship,” Scientific data, 3(1), 2016, pp.1–9. [5] UNESCO, “Recommendation on open educational resources (OER),” Legal Instruments, 2019. [6] L.D. Stein, “Towards a cyberinfrastructure for the biological sciences: progress, visions and challenges,” Nature Reviews Genetics, 9(9), 2008, pp.678–688. [7] N. Wilkins-Diehr, M. Zentner, M. Pierce, M. Dahan, K. Lawrence, L. Hayden, and N. Mullinix, “The science gateways community institute at two years,” In Proceedings of the Practice and Experience on Advanced Research Computing, 2018, pp.1–8. [8] P. Calyam, N. Wilkins-Diehr, M. Miller, E.H. Brookes, R. Arora, A. Chourasia, D.M. Jennewein, V. Nandigam, M. Drew LaMar, S.B. Cleveland, and G. Newman, “Measuring success for a future vision: Defining impact in science gateways/virtual research environments,” Concurrency and Computation: Practice and Experience, 33(19), 2021, p.e6099. [9] S. Donovan and M.D. LaMar, “Using Science Education Gateways to Improve Undergraduate STEM Education: The QUBES Platform as a Case Study,” Computing in Science & Engineering, 25 (2), 2023, pp.20– 29. doi: 10.1109/MCSE.2023.3292313. [10] Gesing S, Lawrence K, Dahan M, Pierce ME, Wilkins-Diehr N, Zentner M. “Science gateways: sustainability via on-campus teams.” Future Generat Comput Syst. 2019;94:97-102. [11] Katz D, Proctor D. “A framework for discussing e-research infrastructure sustainability.” J Open Res Softw. 2014;2(1):e13. [12] C. Clauss, A. Aryal, and K. Ready, “Open Metadata Exchange,” (2025), GitHub repository, https://github.com/ISKME/Open-Metadata-Exchange [13] R. Salz, “InterNetNews: Usenet Transport for Internet Sites.” USENIX Summer, 1992. [14] “InterNetNews,” (2025), GitHub repository, https://github.com/InterNetNews/inn [15] B. Kantor and P. Lapsley. RFC 977 “Network News Transfer Protocol: A Proposed Standard for the Stream-Based Transmission of News,” 1986, https://www.rfc-editor.org/rfc/rfc977.html. [16] S. Ram´ ırez et al., “FastAPI,” (2025), Github repository, https://github.com/fastapi/fastapi. [17] D. LaMar, D. Rook, S. Prescott, and S. Donovan. “Building open education ecosystems to facilitate the adoption and adaptation of computational and data-centric tools across undergraduate biology education,” Open Education Ecosystems RCN, QUBES Educational Resources. doi:10.25334/7QMG-BP02.