scieee AI-readable full text Open interactive document viewer

The Science Explorer and the Future of the Unified Astronomy Thesaurus

Bartlett, Jennifer; Grezes, Felix; Chivvis, Daniel; Alkan, Atilla Kaan; Science Explorer Team

Publisher
Zenodo
DOI
10.5281/zenodo.17221079
Repository
zenodo.org
Language
en

Abstract

The American Astronomical Society (AAS) renewed its commitment to the Unified Astronomy Thesaurus (UAT) by establishing a Working Group on the UAT. The AAS charged them with developing a vision and strategic plan for the UAT over the next five years. When the Wolbach Library closed in March 2024, the UAT lost its half-time curator. Since then, the Working Group has released a new version, 5.1, and is preparing for another by the end of 2025. The next version will expand the number of heliophysics concepts available to better describe research in that discipline. The Astrophysics Data System (ADS) and the Science Explorer (SciX), its multidisciplinary expansion, use the UAT as their preferred keyword system for astronomy, heliophysics, and planetary science. Keyword AI Labeler At Scix (KAILAS) is their developmental AI algorithm that will eventually assign UAT keywords to all space science articles in their collections. Members of the Working Group are providing their technical expertise and evaluating the algorithm’s performance. The UAT is a Simple Knowledge Organization System (SKOS) ontology for the space sciences. The AAS owns the vocabulary but distributes it freely under a liberal Creative Commons license. This open thesaurus is continually growing and adapting to scientific developments through community contributions. In addition to ADS and SciX, the AAS Journals Icarus Publications of the Astronomical Society of the Pacific (PASP) the International Virtual Observatory Alliance (IVOA) Space Telescope Science Institute (STScI) Strasbourg astronomical Data Center (CDS) have adopted the UAT to make articles and data more findable in their systems. When research outputs are more easily found, they are more likely to be cited and re-used, which increases the return on the initial funding investment. The Working Group seeks your feedback on the quality of UAT concepts, the usability of the tools associated with it, and the relevance of KAILAS assignments. They will work with you to address any barriers to your adopting the UAT to describe your data products.

Full text

The Science Explorer and the Future of the Unified Astronomy Thesaurus Jennifer Lynn Bartlett, Felix Grezes, Daniel Chivvis, Attila Alkan, & the SciX Team [email protected] felix.gr[email protected]d.edu [email protected] [email protected] ADEC Meeting, 03 September 2025 This work is openly licensed via CC BY 4.0. Unified Astronomy Thesaurus ●Simple Knowledge Organization System (SKOS) ontology for astronomy ○Owned and freely licensed by the American Astronomical Society (AAS) ○Open, community-contributed resource ○Primarily astronomy concepts, increasing planetary science & heliophysics content ●Keyword system for ○AAS journals ○Publications of the Astronomical Society of the Pacific (PASP) ○Icarus ○International Virtual Observatory Alliance (IVOA) ○Space Telescope Science Institute (STScI) ○Strasbourg astronomical Data Center (CDS) Unified Astronomy Thesaurus https://astrothesaurus.org/ https://github.com/astrothesaurus/UAT 2 Unified Astronomy Thesaurus Intergalactic medium https://astrothesaurus.org/uat/813 Broader Concepts: Extragalactic astronomy Narrower Concepts: Missing mass Quasar absorption line spectroscopy Dark matter Intergalactic clouds Intergalactic filaments Intergalactic gas Intergalactic medium phases Voids Related Concepts: Astrophysical dust processes Intracluster medium Alternate Terms: IGM Intergalactic matter Definition: The matter or environment between the galaxies of a cluster. -An Etymological Dictionary of Astronomy and Astrophysics, by M. Heydari-Malayeri 3 UAT at Crossroads Wolbach Library closing in March 2024 eliminated part-time UAT curator ●previously supported by CfA ●SciX absorbed some technical support responsibilities ●AAS reorganized management of project ●SciX & AAS funded temporary curation contract for knowledge transfer AAS Working Group on UAT 5 years to initiate strategic plan for thesaurus ●Released v. 5.1 summer 2024, next update winter 2025 ●Expanding coverage for heliophysics ●Planning expanded coverage for planetary science ●Investigating alternative curation models ●Modernizing technology infrastructure ●Seeking additional funding sources Daniel Chivvis, member Jennifer Lynn Bartlett, vice chair Nominations open! https://aas.org/comms/working-group-unified-astronomy-thesaurus-wguat 03 UAT for Enhanced Discoverability ●Potential to be single keyword system for space sciences regardless of media ○Consistent, controlled vocabulary for searching ○Describes literature, datasets, and software ○Relates concepts to one another in hierarchy for exploration ●Potential to connect to other ontologies for cross-disciplinary search ●SciX (ADS) has promoted UAT use since 2013 ○Goal: assign UAT keywords to all astronomy articles, including older pre-UAT articles, articles in journals that have not adopted the system, and articles whose authors are parsimonious ○Supplement, not supplant, existing keywords ○Keyword AI Labeler At Scix (KAILAS) is developmental AI algorithm KAILAS: Keyword AI Labeler At Scix KAILAS automatically assigns UAT keywords to astronomy papers. ●KAILAS is a RoBERTa-based model trained on the full text of ~425,00 astronomy papers (3+ million tokens) in our database that have been tagged by authors/publishers with UAT keywords ●The trained model, as well as future improvements, are publicly available: huggingface.co/adsabs/KAILAS ●A version of the dataset with only titles and abstract is publicly available for academic research: huggingface.co/datasets/adsabs/SciX_UAT_keywords ●Development follows publisher agreements mount Kailas from the Himalayas KAILAS Training Input TITLE: Ultrafast Variability in AGN Jets: Intermittency and Lighthouse Effect ABSTRACT: Gamma-ray flares from active galactic nuclei (AGNs) show substantial variability on ultrafast timescales (i.e., shorter than the light-crossing time of the AGN's supermassive black hole). We propose that ultrafast variability is a by-product of the turbulent dissipation of the jet Poynting flux. Due to the intermittency of the turbulent cascade, the dissipation is concentrated in a set of reconnecting current sheets. Electrons energized by reconnection have a strong pitch-angle anisotropy, i.e., their velocity is [...] Output KAILAS ASSIGNED KEYWORDS: Blazars active galactic nuclei Jets relativistic jets plasma astrophysics high energy astrophysics magnetic fields Compare AUTHOR ASSIGNED KEYWORDS: gamma-rays blazars galaxy jets V1 example KAILAS Assessment Feedback from AAS Working Group on UAT ●KAILAS v2 total of 77 papers + 1 control reviewed by 8 experts ●keywords to be relevant 158 times (68%), potentially relevant 36 times and wrong 39 times (17%). ●keywords assigned are varied, mostly correct but not better than human yet ○Author practices erratic ○Uneven coverage ○Poisoned training set -magnitude ! ●More feedback desired!