Controlled Vocabularies and Thesauri in Astronomy: Tracing the Past, Present, and Future

Controlled Vocabularies and Thesauri in Astronomy: Tracing the Past, Present, and Future Rajesh Kumar Das School of Information, College of Communication & Information Florida State University LISA X, 3-7 November 2025 Introduction Controlled vocabularies and thesauri existed long before the advent of computers and were originally designed as independent tools to standardize terminology within specific domains ( Allemang, Hendler, & Gandon, 2020 ). Controlled vocabularies provide a straightforward list of approved terms used for indexing and retrieval. They are foundational tools in systems requiring uniform terminology, such as library catalogs and institutional repositories (Hedden 2010, p. 281). A thesaurus is a type of controlled vocabulary that includes hierarchical, associative, and equivalence relationships between terms. LISA X, 3-7 November 2025 Background •Astronomy is a highly data-intensive field, and controlled vocabularies & thesauri have been playing a vital role in standardizing terminology, enhancing discoverability, and building an astronomical knowledge infrastructure. •They evolved from manual abstracting systems (e.g., AJB, AAA) to digital, AI-enhanced frameworks, such as UAT and SciX. •This study traces the historical evolution, current trends, and future directions of controlled vocabulary and thesaurus-based knowledge organization in astronomy. LISA X, 3-7 November 2025 Research Questions 1. How have controlled vocabularies and thesauri in astronomy evolved from traditional systems to AI-augmented semantic frameworks over time? 2. What trends, strengths, and gaps have emerged from the integration of technology and AI? 3. How can AI-driven and hybrid human–AI approaches shape the future of semantic curation and interoperability? LISA X, 3-7 November 2025 Methodology Approach: Qualitative content analysis conducted via a systematic literature review. Rigor: Adhered to PRISMA 2020 guidelines (Page et al., 2021) to ensure rigor and transparency. Databases: Web of Science (WoS) & SciX. Data Collection: Search was conducted on July 3, 2025, to identify peer-reviewed Englishlanguage articles from all years focusing on broader thesaurus development within the field of astronomy and astrophysics. Analysis: Inductive thematic analysis (Braun & Clarke, 2006). LISA X, 3-7 November 2025 Search Strategy LISA X, 3-7 November 2025 •WoS: TS = (“thesaur*” OR “keyword*” OR “vocabular*” OR “controlled vocabular*”) AND Limits to WoS Subject Category: Astronomy & Astrophysics → 258 •SciX: abs:(“thesaur*” OR “keyword*” OR “vocabular*” OR “controlled vocabular*”) then limits to “Referred” and Collection: Astronomy → 300 Study Selection LISA X, 3-7 November 2025 PRISMA Flowchart for study selection Thematic Overview Five chronological themes were identified as below: Theme Period Focus Representative Citations 1 past–1979 Traditional classification and foundations of astronomical indexing Fricke (1969); AAS (1972); AIP (1975); Schmadel (1979, 1982, 1989); Burkhardt et al. (1989); Shobbrook (1995); Physics Today (2002); Ballmann (2007); IVOA (2007); Frey & Accomazzi (2018); ARIBIB (n.d.) 2 1980–1989 Foundations of controlled vocabularies and thesauri Adorf & Busch (1988); Squibb & Cheung (1988); Egret & Wenger (1988); Schmadel (1989); Shoobrook (1989); Grant et al. (1992); Kurtz et al. (1993); Shobbrook & Shobbrook (1992, 2006); Dorokhova & Dorokhov (2002, 2007); Kurtz et al. (2000) 3 1990–1999 Standardization & interoperable retrieval frameworks Eichhorn (1993); Eichhorn et al. (1995, 1997); Schmitz et al. (1995); Shobbrook (1995); Demleitner et al. (1998); Albrecht & Merkl (1999); Kurtz et al. (1999); Lee et al. (1999); Ortiz et al. (1999); Accomazzi et al. (1995, 1997, 2000); Mothe et al. (2002); Derrière (2003); Eichhorn et al. (2003); Lee & Dubin (2003); Shobbrook & Shobbrook (2006); Ballmann et al. (2007) 4 2000–2009 Emergence of semantic standards and response to fragmentation Murtagh & Guillaume (1998); Accomazzi et al. (2000); Genova et al. (2000); Eichhorn et al. (2003); Derrière et al. (2004); Schwarz (2005); Lesteven et al. (2007); Gray & Ounis (2009); Hill et al. (2009); Hanisch et al. (2015); Chyla et al. (2015); Frey et al. (2018); Grezes et al. (2021); Mampaey et al. (2025) 5 2010–Present Semantic & cognitive shift to AI -ready, community -supported curation Derrière et al. (2004); Gray et al. (2013); Accomazzi et al. (2014); Frey et al. (2014); Frey & Accomazzi (2018); Grezes et al. (2021, 2022a, 2022b, 2024); Hurlburt & Timmons (2022); Shapurian et al. (2023); Poduval et al. (2023); Gray et al. (2023); Blanco-Cuaresma et al. (2024); Lockhart et al. (2024); Iyer et al. (2024); Accomazzi (2024); Cecconi et al. (2025); Bartlett et al. (2025) LISA X, 3-7 November 2025 Theme 1: Traditional Classification and the Foundations of Astronomical Indexing (Past–1979) •Launch of Astronomischer Jahresbericht (AJB) in Vol. 7; 1st systematic classification for astronomy (1905); later used by Astronomy and Astrophysics Abstracts (AAA) with a hierarchical taxonomy •Building of conceptual foundation for controlled vocabularies in astronomy •AAA launched (1969) under the IAU, succeeding AJB •Introduced scientific keyword assignment •Physics and Astronomy Classification Scheme by the American Institute of Physics (1975) •Introduced a hierarchical system of numeric codes for physics and astronomy and widely adopted in major scientific journals and indexing services •Various subject indexes and keyword lists were developed by independent journals (e.g., ApJ, A&A) LISA X, 3-7 November 2025 Theme 4: Gaps or Challenges •The absence of a unified metadata framework led projects such as SOLARNET VO to favor flexibility over strict standardization. •OCR-based text extraction in ADS introduced frequent recognition errors, necessitating extensive manual correction and validation. •UCDs standardized terminology but lacked an ontological framework to capture hierarchical or logical relationships. •Linking DOIs to specific dataset queries or metadata selections was unsupported, requiring new modules for precise and citable data connections. LISA X, 3-7 November 2025 Theme 5: The Semantic and Cognitive Shift: AI Integration and FAIR Data (2010–Present) •Unified Astronomy Thesaurus (UAT) was initiated (2012) after PACS decommissioning to merge IAU, PACS, IVOAT, and ASK vocabularies. Released in 2013; stewarded by AAS + ADS + IVOA. •UAT adopted SKOS/OWL alignment, IVOA URIs, and FAIR-compliant interoperability; maintained in SKOS/RDF for Linked Data; used a hybrid workflow: community GitHub suggestions → editorial review → librarian approval; enabled AI-based keyword assignment within ADS •Planetary Data System set an ontology-based standard for metadata semantics; FAIR-IMPACT project (2022–25) and OntoPortal-Astro initiative build a shared registry of Semantic Artefacts (SAs). •astroBERT and AstroLLaMA LLMs enable semantic search and question answering; Pathfinder (ADS) implements Retrieval-Augmented Generation (RAG) for factgrounded answers; ML visualizations identify concept gaps in the UAT. •SciX broadens ADS to all five NASA SMD disciplines; KAILAS automates metadata enrichment, UAT keyword assignment, and planetary feature tagging. LISA X, 3-7 November 2025 Theme 5: Gaps or Challenges •Preparing legacy data for AI remains slow and costly, making full AI-readiness within a decade unrealistic. •Incompatible XML or text-based vocabularies hinder cross-domain FAIR data alignment. •LLMs can produce convincing but inaccurate outputs, requiring source verification. •Linking papers to datasets still needs manual checks due to limited SPASE metadata access. •The UAT lacks full coverage of emerging computational and observational concepts. •LLM tools like Pathfinder rely on abstracts, missing deeper methodological details. LISA X, 3-7 November 2025 Future Research Agenda •Enhance LLM-powered retrieval using Retrieval-Augmented Generation (RAG) to improve ADS/SciX search precision and minimize hallucinations. •Develop AI-driven user interfaces such as SciX Chat Assistants and API tools for conversational, intelligent data exploration. •Use AI and LLMs to automate schema mapping and vocabulary alignment across diverse Semantic Artefacts (SAs). •Integrate LLMs with knowledge graphs and symbolic reasoning to refine semantic retrieval and query expansion. •Employ multimodal AI systems like Pathfinder, AstroAI, and UniverseTBD to enable automated hypothesis generation and scientific discovery. LISA X, 3-7 November 2025 Limitations •The review was confined to peer-reviewed English-language papers in WoS and SciX, excluding grey literature and nonEnglish sources. •Thematic analysis depended on available documentation, which may have overlooked emerging or unpublished developments. LISA X, 3-7 November 2025 Conclusion •Astronomical knowledge organization has evolved from manual cataloging (AJB, AAA) → interoperable standard (IVOA) → AI-augmented infrastructures (UAT, SciX). •The field now embraces ontology-driven, community-supported, and FAIRaligned systems, enhancing scalability, consistency, and discovery. •The rise of AI as a co-curator and co-discoverer marks a paradigm shift in metadata enrichment, vocabulary alignment, and hypothesis generation. •Yet, AI-readiness requiring standardized metadata, documentation, and FAIR compliance remains a pressing challenge in astronomy/ space science and neighboring field-related data. LISA X, 3-7 November 2025 References •Allemang, D., Hendler, J., & Gandon, F. (2020). Semantic web for the working ontologist: Effective modeling for linked data and semantic web (3rd ed.). Morgan Kaufmann. •Braun, V., & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research in Psychology, 3(2), 77–101. https://doi.org/10.1191/1478088706qp063oa •Hedden, H. (2010). The accidental taxonomist. Information Today, Inc. •Page, M. J., McKenzie, J. E., Bossuyt, P. M., Boutron, I., Hoffmann, T. C., Mulrow, C. D., ... Moher, D. (2021). The PRISMA 2020 statement: An updated guideline for reporting systematic reviews. BMJ, 372, n71. https://doi.org/10.1136/bmj.n71 LISA X, 3-7 November 2025 Acknowledgments This research was conducted as part of the “Research Collaboration” project at Florida State University, supervised by: •Dr. Gretchen Stahlman Assistant Professor College of Communication and Information School of Information Florida State University •LISA 10 OC LISA X, 3-7 November 2025 Questions!! Email: [email protected] LISA X, 3-7 November 2025