Taming the Space: Semantic Modelling using Linked Open Data of Fuzzy Wobbling Geospatial Data from the Archaeological and Geosciences domains

Taming the Space: Semantic Modelling using Linked Open Data of Fuzzy Wobbling Geospatial Data from the Archaeological and Geosciences domains Florian Thiery* 1,2 , Fiona Schenk 3,2 , Stefanie Baars 4 , Allard W. Mees 1 , Sophie C. Schmidt 5,2 , Chiara G.M. Girotto 6 , Fabian Fricke 7 1 Leibniz-Zentrum für Archäologie (LEIZA) – Mainz, Germany 2 Research Squirrel Engineers Network – Mainz, Germany 3 Johannes Gutenberg Universität Mainz (JGU) – Mainz, Germany 4 Münzkabinett, Staatliche Museen zu Berlin, SPK – Berlin, Germany 5 Free University of Berlin – Berlin, Germany 6 Archaeobiocenter | LMU Munich – Munich, Germany 7 German Archaeological Institute (DAI) – Mainz, Germany * Corresponding author Correspondence: fl[email protected] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 A BSTRACT Archaeological and geoscientific research often relies on spatial data whose precision and reliability vary considerably. Coordinates derived from excavation reports, literature, or modern surveys frequently embody vagueness, ambiguity, or conflicting interpretations. This paper addresses such wobbling geospatial data by presenting a semantic approach that embeds uncertainty directly into Linked Open Data (LOD). At the core of this approach lies the Fuzzy Spatial Locations Ontology (FSL-O), which builds on PROV-O, SKOS, and GeoSPARQL to model entities, activities, and agents alongside explicit properties for certainty, provenance, and methodological detail. Its implementation in the fuzzy-sl Wikibase extends this framework through a dedicated schema for geolocation and coordinate metadata, enabling the documentation of precision, methods, sources, and actors. The SPARQL Unicorn Toolkit further supports visualisation by integrating these data into GIS environments and generating human-readable HTML outputs. The methodology is demonstrated through interdisciplinary case studies: Ogham stones with multiple biographies between ancient findspots and museum contexts; Campanian Ignimbrite deposits serving as Late Pleistocene chronological markers; poorly documented numismatic hoards from Croton; and Roman Samian ware distributions. Each example illustrates how uncertainty can be expressed as structured, machine-readable information, rather than being hidden or oversimplified. By embedding fuzziness into LOD, this work enhances the FAIRness, transparency, and interoperability of archaeological and geoscientific datasets. It contributes to international debates on how to responsibly represent spatial uncertainty and provides a foundation for future extensions of ontology and Wikibase models across broader cultural heritage and scientific domains. Keywords: FAIR, Linked Open Data, uncertainty, Archaeology, Ogham, Samian Ware, Geosciences, Campanian Ignimbrite, Wikibase 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 Introduction & Background Archaeological and geoscientific research is deeply reliant on spatial data, yet the production and use of such data are rarely free from uncertainty [1], [2]. Coordinates derived from historical sources, excavation reports, field surveys, or secondary literature often contain vagueness, ambiguities, or imprecision that complicate their interpretation and reuse [3], [4], [5], [6]. This situation gives rise to what we refer to as “fuzzy wobbling geospatial data”, locations that shift depending on the source, method, or interpretative framework employed [7]. Unless such uncertainty is explicitly recorded, the resulting data cannot be considered transparent, reproducible, or reusable in the long term. This challenge aligns directly with the FAIR Principles [8], now widely recognised as a cornerstone of modern research data management. Achieving FAIRness requires more than simply publishing datasets: the context of their creation, including the methods and assumptions underlying geospatial coordinates, must also be documented. For archaeological and geoscientific datasets, where site locations may be vague or disputed, this transparency is essential to ensure trust in the data and to enable meaningful reuse across disciplinary boundaries. Figure 1 - A distributed Knowledge Graph Scheme bringing together Linked Open Data and FAIR Digital Objects approaches. Florian Thiery & Andreas Noback, CC BY 4.0, via Wikimedia Commons. Linked Open Data (LOD) [9], [10], [11] offers a framework for addressing these requirements. By representing information in graph-based structures using standards such as RDF, SPARQL, and GeoSPARQL [12], [13], LOD allows heterogeneous datasets to be semantically connected and enriched. This approach has particular relevance for archaeology, a discipline that inherently integrates diverse forms of evidence, from geological stratigraphy and palaeoenvironmental records to artefact typologies and textual sources. LOD enables the integration of such evidence within the federated 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 interdisciplinary knowledge graph (FIKG, Fig. 1), while at the same time enabling the explicit modelling of uncertainty and provenance. The need for explicit modelling is especially evident when spatial information is transferred between communities and infrastructures. Geoscientists may describe a volcanic eruption layer in stratigraphic terms. At the same time, archaeologists may refer to associated findspots, and historians may only mention the nearest settlement without a shared semantic model that can capture the varying degrees of certainty, precision, and spatial reference; such information risks remaining siloed, ambiguous, or even misleading. By contrast, modelling uncertainty as part of the dataset itself enables a richer and more honest representation of the past. In recent years, several initiatives have highlighted the importance of documenting vagueness and uncertainty in geospatial data, particularly through semantic technologies. The adoption of ontology-driven approaches allows doubts and imprecisions to be represented in a structured, machine-actionable way, rather than being relegated to footnotes or omitted entirely. This ensures that subsequent users – whether domain experts, interdisciplinary collaborators, or computational systems – can interpret the confidence, methods, and limitations associated with each spatial statement. This paper builds on these developments by presenting a semantic modelling approach for wobbling geospatial data. We focus on the Fuzzy Spatial Locations Ontology (FSL-O), designed to capture vagueness, uncertainty, and provenance in georeferenced data. Implemented within a federated ecosystem of Wikidata and dedicated Wikibase instances, FSL-O provides a framework for integrating archaeological and geoscientific datasets while making explicit the circumstances of their creation. The paper further demonstrates how this model can be applied in practice through a series of interdisciplinary case studies, ranging from early medieval Ogham stones to Late Pleistocene volcanic deposits and ancient numismatic hoards. By foregrounding uncertainty as an integral component of data modelling, our approach seeks not only to enhance the FAIRness of archaeological and geoscientific data but also to provide a methodological foundation for their long-term interoperability. In doing so, we aim to contribute to a broader discussion within the CAA community on how best to represent, integrate, and reuse spatial data that, by its very nature, continues to wobble. Methodology: SPARQL Unicorn & fuzzy-sl Ontology The methodological core of this paper lies in the semantic modelling of vagueness and uncertainty in geospatial data. While spatial coordinates are typically treated as fixed and precise points in databases, archaeological and geoscientific practice reveals a far more complex reality. Locations may be derived from excavation reports, secondary literature, oral traditions, or modern field surveys, each with varying degrees of precision and confidence. To make such variability transparent and reusable, 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 uncertainty itself must become part of the data model. In the following, we present the main tools and ontological frameworks that enable such modelling and visualisation: the SPARQL Unicorn Toolkit, the Fuzzy Spatial Locations Ontology (FSL-O), and the fuzzy-sl Wikibase. Figure 2 - Scheme of “fuzzy-sl” modelling, based on PROV-O, with the classes Entity, Activity and Agent. Florian Thiery, CC BY 4.0. The SPARQL Unicorn Toolkit [14], [15], [16], [17] provides the technical infrastructure for visualising Linked Open Data in GIS and for transforming machine-actionable RDF into human-readable outputs. Originally conceived as a broader FAIRification tool, in this paper, its role is restricted to two key functions: the integration of LOD into GIS environments such as QGIS and the generation of an LOD HTML documentation. Through its QGIS plugin, Linked Data queries can be performed directly against SPARQL endpoints, enabling the overlay of uncertain or fuzzy locations with other spatial datasets. For example, queries for Ogham stones, Campanian Ignimbrite findspots, or numismatic hoards can be executed and displayed as clustered layers in QGIS. The 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 same toolchain also enables the creation of HTML pages that render RDF triples in a navigable, human-readable form. In this way, the Unicorn Toolkit acts as an interface layer between the technical modelling of uncertainty and its use in everyday research practice. Figure 3 - fuzzy-sl Wikibase modelling scheme. Florian Thiery, CC BY 4.0. At the centre of this work, however, is the Fuzzy Spatial Locations Ontology (FSL-O), a semantic model specifically designed to capture what we refer to as “wobbling geospatial data.” FSL-O [18], [1] is built upon three established standards: PROV-O, SKOS, and GeoSPARQL ( Fig. 2 ). Following the PROV-O concept [19], it distinguishes between entities , activities , and agents . A site is represented as an entity associated with a geometry, which is created by a method understood as an activity, and this activity is attributed to a person, the agent. This triadic structure ensures that every coordinate can be traced back to the procedure and authority by which it was produced. Crucially, FSL-O incorporates explicit properties to record uncertainty. Each site and its geometry can be described by two core attributes: fsl:certaintyDesc , a textual description of confidence or doubt, and fsl:certaintyLevel , a controlled vocabulary ranging from low to high, including dubious. Sites may further be linked to bibliographic references via fsl:hasReference or to online resources through exactMatch properties borrowed from the SKOS vocabulary. Methods, in turn, can be characterised through fsl:hasSource and fsl:hasSourceType , which identify the underlying data source and its category, fsl:activityDesc , which 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 documents the method in free text, and the already mentioned certainty properties, which carry over to the level of activities. Together, these properties allow not only the representation of where a site is located, but also the circumstances, evidence, and confidence with which this location has been established. The resulting RDF can then be transformed into HTML files through the SPARQL Unicorn Ontology Documentation research tool, ensuring that the ontology-driven data remains both machine-processable and accessible to human users. Figure 4 - Example Wikibase "fuzzy-sl" modelling of Ai/Aii: Ogham Stone Coumeenole North (Q131) with current location (Ai) and finding site (Aii); B: Campanian Ignimbrite Findspot Auel Maar AU3 (Q70); C: Campanian Ignimbrite Findspot Urluia Quarry (Q73). Florian Thiery, CC BY 4.0. While FSL-O provides the conceptual framework, its integration into existing community infrastructures is essential. Wikidata, as one of the most widely used community-driven knowledge graphs, allows the modelling of spatial information via the property P625 for geographic coordinates. However, such coordinates can be enriched with qualifiers and references to capture uncertainty. For instance, the modelling of an Ogham stone in Wikidata can distinguish between its ancient findspot and its modern exhibition location. Qualifiers such as determination method (P459), subject has role (P2868), and stated in (P248) enable the documentation of whether a coordinate was derived from georeferencing, obtained from a database entry, or observed on-site. The sourcing circumstances (P1480) qualifier can indicate whether the information was obtained through a direct survey. At the same time, location (P276) distinguishes between the different roles a place may play, such as a findspot, an exhibition venue, or a general reference. References to external resources, such as OpenStreetMap node identifiers (P11693), provide additional anchoring of the coordinate to collaborative geospatial repositories. By combining these qualifiers and references, a Wikidata item 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 can reflect both the ancient and modern biographies of a site, while making explicit the methods and uncertainties associated with each location. The most detailed implementation of the ontology-driven approach is achieved through the fuzzy-sl Wikibase, a specialised Wikibase instance dedicated to modelling uncertain geospatial data (Fig. 3). Its data model distinguishes between Geolocation Metadata (GLM), represented as statements, and Coordinate Metadata (CM), represented as qualifiers. GLM includes properties such as relatedTo (P10), which connects a site to external resources in the LOD cloud; hasReference (P11), which records a literature quotation or bibliographic source; hasSpatialType (P8), which specifies categories like inhabited place, river, cave, maar, or supervolcano; hasSpatialCategory (P26), which situates the location within domains such as geology or archaeology; and locatedInTheAdministrativeTerritorialEntity (P32), which links to corresponding Wikidata items for administrative units. These high-level statements establish the basic context of a site. The current ten CM qualifiers then provide fine-grained metadata about the coordinates themselves. Precision (P23) records the number of significant decimal places in the WGS84 coordinate, thereby expressing the numerical accuracy of the georeferencing (for example, ±0.0001° for the Franchthi Cave). CertaintyLevel (P5) and certaintyDescription (P13) mirror the FSL-O properties, allowing both a categorical and textual account of confidence. MethodUsed (P7) and methodDescription (P15) document how the coordinate was obtained, whether by external repository, literature-based georeferencing, or on-site survey. ActingPerson (P14) identifies the individual responsible for the georeferencing activity, while sourceTypeGeneric (P6) and sourceTypeDetail (P16) provide further differentiation of the evidence base, distinguishing between textual descriptions, papers, or community-contributed sources such as OpenStreetMap. Finally, locationType (P24) specifies the functional category of the location – whether it is a findspot, eruption site, or museum display – while pointType (P33) indicates the conceptual role of the coordinate, such as a representative point, a place as a concept, or a natural place. By combining GLM and CM, the fuzzy-sl Wikibase thus enables the explicit recording of both the contextual and methodological dimensions of geospatial data (Fig. 4). The interplay of these three tools—SPARQL Unicorn for visualisation, FSL-O for conceptual modelling, and the fuzzy-sl Wikibase for implementation – creates a workflow that renders geospatial uncertainty transparent, reusable, and interoperable. Rather than presenting a fixed or oversimplified representation of the past, this approach acknowledges the wobbling nature of spatial data. It provides the means to document, visualise, and critically engage with it. The following case studies will demonstrate how this methodology can be applied to diverse datasets, ranging from epigraphic monuments and volcanic tephra layers to numismatic hoards and ceramic distributions. Case Studies: Ogham, Campanian Ignimbrite & Samian Ware The methodological approaches outlined above gain their full relevance only when applied to concrete research data. In the following, we demonstrate how the modelling of fuzzy and wobbling geospatial data can be put into practice across a selection of 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 interdisciplinary case studies. The examples highlight how FSL-O and the fuzzy-sl Wikibase allow for a transparent representation of uncertainty, how the SPARQL Unicorn Toolkit facilitates visualisation, and how Linked Open Data infrastructures enable integration across domains. Figure 5 - Geospatial modelling of Ogham Stone CIIC 81 / Q74 / GARES/1 [20] at UCC Stone Corridor, no. IV. Florian Thiery, CC BY 4.0. One of the longest-running and most illustrative examples is the Linked Open Ogham project. Ogham stones, erected mainly between the fourth and seventh centuries CE in Ireland and parts of Britain, are among the earliest written monuments of the Irish language [21], [22], [23]. They provide not only linguistic evidence but also insights into kinship structures, territorial organisation, and cultural contacts. Many stones, however, have complex biographies: some remain in situ, others were moved into museums, and many are known only from antiquarian reports. The fuzziness of their spatial references is thus intrinsic. Using the fuzzy-sl Wikibase, Ogham stones can be modelled with coordinates (P4) enriched by qualifiers and references. For instance, the stone CIIC 81 (Q74 1 ), initially found at Garranes in County Cork, is documented both in its ancient findspot and in its current location in the UCC Stone Corridor no. 4 (Fig. 5). The exhibition site respectively findspot can be qualified as P5: High / Low; P13: on-site survey at exhibition area / GARES/1 [20]; P7: on-site survey / Georeferencing; P14: Florian Thiery; P6: on-site survey / Textual Description; P16: on-site survey / Paper; P15: on-site survey at UCC Cork / found in old documents; P23: ±0.0001°; P24: Exhibition Site / Findspot; P33: Exhibition / Location of a Place as a Concept. CIIC 81, respectively, GARES/1 is also referenced to other resources in the FIKG, such as Wikidata (Q130529871), Semantic Kompakkt (Q1262), FactGrid (Q10000294), OSM node 11071361392, and SMR ID CO074-148----. This dual modelling captures the spatial biography of the stone and the different confidence levels associated with each coordinate. In practice, these statements can be queried via SPARQL and visualised in QGIS through the Unicorn plugin, which enables both density maps and individual feature exploration. The Ogham case demonstrates how doubt, movement, and 1 cf. https://fuzzy-sl.wikibase.cloud/entity/Q74 (all links last accessed 09.09.2025) 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 Discussion & Outlook The results presented in this paper demonstrate how semantic modelling can transform the way archaeological and geoscientific datasets are documented, compared, and integrated. The focus here has been less on introducing the general problem of uncertainty – already well recognised in both fields—and more on showing how Linked Open Data provides the methodological means to deal with it in a structured and interoperable manner. By applying the Fuzzy Spatial Locations Ontology (FSL-O) and implementing it in the fuzzy-sl Wikibase, combined with visualisation through the SPARQL Unicorn Toolkit, we have highlighted a workflow that captures not only coordinates but also their provenance, confidence, and context. The key methodological contribution of this work lies in bringing together three elements: first, a formal ontology (FSL-O) that expresses vagueness and fuzziness through explicit properties such as certainty level, certainty description, and references to sources; second, a Wikibase implementation that extends this model through a detailed schema for geolocation and coordinate metadata, allowing users to record precision, methods, and agents; and third, an integration pathway to community-driven infrastructures like Wikidata and OpenStreetMap, where data can be connected across disciplinary and linguistic boundaries. This combination moves beyond traditional gazetteers or databases by embedding uncertainty directly into the data model rather than treating it as an external annotation. The case studies underline the value of this approach. Ogham stones exemplify the complex biographies of archaeological monuments, as they shift between findspots and museum contexts. Campanian Ignimbrite deposits demonstrate how geoscientific and archaeological data can be integrated when their uncertainties are modelled explicitly. Numismatic hoards from Croton highlight the varying granularity of spatial descriptions, from precise urban locations to broad regional references. Samian sites illustrate how even well-established research databases require careful treatment of coordinate quality and precision. In all these cases, semantic modelling enables researchers to express differences in confidence, to trace methods of data creation, and to connect to external resources. The interoperability of these datasets is enhanced precisely because uncertainty is preserved as structured information. The integration of the archaeometallic, ceramic, and cultural-geographical case studies further illustrates how the fuzzy-sl approach transcends disciplinary boundaries. By applying the same semantic logic to data originating from metallurgy, regional archaeology, and landscape research, these examples demonstrate that uncertainty is not confined to a single data type or scale. Instead, it manifests as a universal property of archaeological evidence, which can be formally represented and computationally analysed through shared ontological structures. This reinforces the potential of the fuzzy-sl framework to act as a methodological bridge between site-based documentation, regional synthesis, and cross-domain data infrastructures. What sets this paper apart methodologically is its emphasis on Linked Open Data as the backbone of interoperability. The use of RDF, PROV-O, SKOS, and GeoSPARQL 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 ensures that the data produced can be queried, linked, and visualised across infrastructures. The explicit inclusion of uncertainty into this semantic framework is what makes it distinctive. Rather than offering certainty where there is none, the approach allows data to remain honest about its limitations while still being reusable. This is particularly important in international and interdisciplinary contexts, where researchers from archaeology, history, and the geosciences must be able to evaluate not only the data itself but also the degree of confidence attached to it. Visualisation is another critical area for discussion. While modelling uncertainty semantically provides the foundation, researchers also need intuitive ways to see and interpret it. The use of bounding boxes, buffers, or hulls makes fuzziness visible in maps, but these techniques require careful calibration. For example, buffers generated in WGS84 coordinates risk producing distortions that are technically correct but visually misleading. The SPARQL Unicorn Toolkit addresses part of this challenge by enabling semantically annotated data to be pulled into GIS environments where such visualisations can be tested, compared, and refined. In this sense, the toolkit provides not just a display mechanism but also an experimental environment for developing user-friendly representations of uncertain space. Looking forward, the next steps involve extending the ontology and the Wikibase schema in response to further case studies. While the current model already incorporates ten key properties for coordinate metadata, additional attributes may be needed to capture temporal vagueness, different types of spatial abstraction, or more nuanced descriptions of methods. Applying the model to new datasets—such as archaeometallurgical evidence, large-scale cultural landscapes, or additional ceramic corpora—will help to identify such requirements. The modularity of FSL-O and the flexibility of Wikibase make them well-suited for this iterative refinement. Another avenue for development lies in alignment with international standards and infrastructures. Mapping FSL-O to CIDOC CRM and its archaeological extensions, or linking Wikibase data to global initiatives such as the European Collaborative Cloud for Cultural Heritage (ECCCH), will ensure that fuzzy geospatial data can participate in wider ecosystems. This is particularly important for long-term sustainability: only if data models are compatible across infrastructures can they be maintained and reused at scale. In conclusion, the central contribution of this paper is methodological. By embedding uncertainty into Linked Open Data through the combined use of FSL-O, fuzzy-sl Wikibase, and SPARQL Unicorn visualisation, we have shown a concrete pathway for dealing with wobbling geospatial data in archaeology and the geosciences. The approach is internationally relevant, interdisciplinary by design, and scalable through alignment with broader infrastructures. Future work will focus on expanding the ontology and property sets, integrating further case studies, and refining visualisation strategies. In this way, the model not only acknowledges the inherent fuzziness of spatial information but also provides the tools to make it transparent, interoperable, and FAIR. 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 Acknowledgements The authors would like to thank Karsten Tolle and Sebastian Gampe for their advice, as well as the CAA Germany, SIG Data Dragon and NFDI4Objects Community. The authors acknowledge the use of language assistance powered by artificial intelligence (ChatGPT, OpenAI) for stylistic editing and linguistic refinement. All content and arguments were authored and verified by the authors themselves. Data, scripts, code, and supplementary information availability Thiery, F. et al. (2025). fuzzy-sl Wikibase: https://fuzzy-sl.wikibase.cloud ; Thiery, F., & Schenk, F. (2023). Campanian Ignimbrite Geo Locations [DataSet] at https://github.com/Research-Squirrel-Engineers/campanian-ignimbrite-geo [49]; Thiery, F., & Baars, S. (2023). Croton Geo Locations [DataSet] at https://github.com/Research-Squirrel-Engineers/croton-geo ; Thiery, F. (2023). Fuzzy Spatial Locations Ontology [DataSet] at https://github.com/Research-Squirrel-Engineers/fuzzy-sl-ontology [50]; Thiery, F., & Homburg, T. (2025). SPARQLing Unicorn QGIS Plugin [Software] at https://github.com/sparqlunicorn/sparqlunicornGoesGIS [14], [15], [16], [17]; Homburg, T., & Thiery, F. (2024). SPARQL Unicorn Ontology Documentation [Software] at https://github.com/sparqlunicorn/sparqlunicornGoesGIS-ontdoc [14], [15], [16], [17]. Conflict of interest disclosure The authors declare that they comply with the PCI rule of having no financial conflicts of interest in relation to the content of the article. Funding This paper is part of the DFG-funded NFDI initiative, specifically the Research Data Infrastructure for the Material Remains of Human History (NFDI4Objects), DFG Project number 501836407. Some works were funded by Wikimedia Deutschland, FellowProgramm Freies Wissen 2020/21: “Irische Ogham Steine im Wikimedia Universum”. References [1] F. Thiery, F. Schenk, S. Baars, K. Tolle, and P. Thiery, ‘Modellierung von Fuzzyness / Wobbliness in Geodaten - Am Beispiel archäologischer und geowissenschaftlicher Fundortreferenzen’, Tagungsband FOSSGIS-Konferenz 2024 , vol. 2024, pp. 64–73, Mar. 2024, doi: 10.5281/zenodo.10571858. [2] N. J. Car, ‘Spatial Uncertainty for Features & Functions (SUFF) Model’, Spatial Uncertainty for Features & Functions (SUFF) Model. Accessed: Oct. 21, 2024. [Online]. Available: https://w3id.org/suff/spec [3] F. Thiery, F. Schenk, and S. Baars, ‘Dealing with doubts: site georeferencing in archaeology and in the geosciences’, Archeologia e Calcolatori , vol. 35, no. 2, pp. 97–106, 2024, doi: 10.19282/ac.35.2.2024.11. 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 [4] F. Thiery and F. Schenk, ‘Modelling of Uncertainty in Geo Sciences Sites’, Squirrel Papers , vol. 5, no. 1, p. #4, Dec. 2023, doi: 10.5281/zenodo.10255259. [5] F. Thiery and F. Schenk, ‘How to locate the Campanian Ignimbrite site Urluia based on literature? How to provide and publish this data in a FAIR way?’, Squirrel Papers , vol. 5, no. 1, p. #5, 2023, doi: 10.5281/zenodo.10262720. [6] F. Thiery, F. Schenk, and S. Baars, ‘Dealing with doubts: Site georeferencing in archaeology and in the geosciences’, Squirrel Papers , vol. 5, no. 1, p. #6, Dec. 2023, doi: 10.5281/ZENODO.10291889. [7] A. Züfle et al. , ‘Handling Uncertainty in Geo-Spatial Data’, in 2017 IEEE 33rd International Conference on Data Engineering (ICDE) , Apr. 2017, pp. 1467–1470. doi: 10.1109/ICDE.2017.212. [8] M. D. Wilkinson, M. Dumontier, Ij. J. Aalbersberg, G. Appleton, and et al., ‘The FAIR Guiding Principles for scientific data management and stewardship’, Scientific Data , vol. 3, p. 160018, Mar. 2016, doi: 10.1038/sdata.2016.18. [9] S. C. Schmidt, F. Thiery, and M. Trognitz, ‘Practices of Linked Open Data in Archaeology and Their Realisation in Wikidata’, Digital , vol. 2, no. 3, pp. 333–364, June 2022, doi: 10.3390/digital2030019. [10] T. Berners-Lee, ‘Linked Data’. Accessed: May 31, 2024. [Online]. Available: https://www.w3.org/DesignIssues/LinkedData.html [11] F. Thiery and P. Thiery, ‘Linked Open Ogham. How to publish and interlink various Ogham Data?’, Archeologia e Calcolatori , vol. 34, no. 1, pp. 105–114, 2023, doi: 10.19282/ac.34.1.2023.12. [12] N. J. Car and T. Homburg, ‘GeoSPARQL 1.1: Motivations, Details and Applications of the Decadal Update to the Most Important Geospatial LOD Standard’, IJGI , vol. 11, no. 2, p. 117, Feb. 2022, doi: 10.3390/ijgi11020117. [13] J. Abhayaratna et al. , ‘OGC Benefits of Representing Spatial Data Using Semantic and Graph Technologies’, OGC White Paper . OGC, Oct. 05, 2020. [Online]. Available: http://www.opengis.net/doc/wp/using-semantic-graph [14] F. Thiery, F. Schenk, and P. Thiery, ‘Das Research Squirrel Engineers Network: FAIRification Tools und LOD-Projekte aus der Archäoinformatik und den Geowissenschaften’, Archäologische Informationen , vol. 47, no. NWDVA 2024 Bochum: Digitale Archäologie, pp. 115–140, Apr. 2025, doi: 10.11588/ai.2024.1.110394. [15] F. Thiery and T. Homburg, ‘SPARQLing Unicorn QGIS Plugin’, Squirrel Papers , vol. 6, no. 2, p. #5, Sept. 2024, doi: 10.5281/zenodo.13828632. [16] T. Homburg and F. Thiery, ‘SPARQL Unicorn Ontology Documentation’, Squirrel Papers , vol. 6, no. 2, p. #2, Mar. 2024, doi: 10.5281/zenodo.10780476. [17] T. Homburg and F. Thiery, ‘The SPARQL Unicorn Ontology documentation: Exposing RDF geodata using static GeoAPIs’, Tagungsband FOSSGIS-Konferenz 2024 , vol. 2024, pp. 82–90, Mar. 2024, doi: 10.5281/zenodo.10570985. [18] F. Thiery, ‘Fuzzy Spatial Locations Ontology’, Squirrel Papers , vol. 5, no. 2, p. #3, Dec. 2023, doi: 10.5281/zenodo.10362777. [19] ‘PROV-O: The PROV Ontology’. Accessed: Dec. 05, 2023. [Online]. Available: https://www.w3.org/TR/prov-o/ [20] CISP Database, ‘GARES/1 [CISP Database]’. Accessed: Sept. 09, 2025. [Online]. Available: https://linkedopenogham.github.io/cisp-recovery/cisp/database/stone/gares_1.html [21] R. A. S. Macalister, Corpus inscriptionum insularum Celticarum . Dublin: Stationery Office, 1945. 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 [22] D. MacManus, A Guide to Ogam . in Maynooth monographs, no. 4. Maynooth: An Sagart, 1997. [23] S. C. Schmidt and F. Thiery, ‘SPARQLing Ogham Stones: New Options for Analyzing Analog Editions by Digitization in Wikidata’, CEUR Workshop Proceedings , vol. 3110, no. Graph Technologies in the Humanities 2020, pp. 211–244, Mar. 2022, doi: 10.5281/zenodo.6380914. [24] F. Schenk, U. Hambach, S. Britzius, D. Veres, and F. Sirocko, ‘A Cryptotephra Layer in Sediments of an Infilled Maar Lake from the Eifel (Germany): First Evidence of Campanian Ignimbrite Ash Airfall in Central Europe’, Quaternary , vol. 7, no. 2, p. 17, Mar. 2024, doi: 10.3390/quat7020017. [25] B. Giaccio et al. , ‘The Campanian Ignimbrite and Codola tephra layers: Two temporal/stratigraphic markers for the Early Upper Palaeolithic in southern Italy and eastern Europe’, Journal of Volcanology and Geothermal Research , vol. 177, no. 1, pp. 208–226, Oct. 2008, doi: 10.1016/j.jvolgeores.2007.10.007. [26] F. G. Fedele, B. Giaccio, and I. Hajdas, ‘Timescales and cultural process at 40,000 BP in the light of the Campanian Ignimbrite eruption, Western Eurasia’, Journal of Human Evolution , vol. 55, no. 5, pp. 834–857, Nov. 2008, doi: 10.1016/j.jhevol.2008.08.012. [27] M. W. Morley and J. C. Woodward, ‘The Campanian Ignimbrite (Y5) tephra at Crvena Stijena Rockshelter, Montenegro’, Quat. res. , vol. 75, no. 3, pp. 683–696, May 2011, doi: 10.1016/j.yqres.2011.02.005. [28] T. Tsanova et al. , ‘Upper Palaeolithic layers and Campanian Ignimbrite/Y-5 tephra in Toplitsa cave, Northern Bulgaria’, Journal of Archaeological Science: Reports , vol. 37, p. 102912, June 2021, doi: 10.1016/j.jasrep.2021.102912. [29] K. E. Fitzsimmons and U. Hambach, ‘Loess accumulation during the last glacial maximum: Evidence from Urluia, southeastern Romania’, Quaternary International , vol. 334–335, pp. 74–85, June 2014, doi: 10.1016/j.quaint.2013.08.005. [30] S. Pötter et al. , ‘Disentangling Sedimentary Pathways for the Pleniglacial Lower Danube Loess Based on Geochemical Signatures’, Front. Earth Sci. , vol. 9, p. 600010, Apr. 2021, doi: 10.3389/feart.2021.600010. [31] F. Thiery, A. W. Mees, and J. B. Kiesling, ‘Challenges in research community building: integrating Terra Sigillata (Samian) research into the Wikidata community’, AeC , vol. 34, no. 1, pp. 157–164, 2023, doi: 10.19282/ac.34.1.2023.17. [32] M. Reddé and A. Mees, ‘Hadrian’s Wall and its Continental Hinterland’, Britannia , vol. 53, pp. 55–84, Nov. 2022, doi: 10.1017/S0068113X22000216. [33] M. Flückiger, E. Hornung, M. Larch, M. Ludwig, and A. Mees, ‘Roman Transport Network Connectivity and Economic Integration’, The Review of Economic Studies , vol. 89, no. 2, pp. 774–810, Mar. 2022, doi: 10.1093/restud/rdab036. [34] R. Grünthal et al. , ‘Drastic demographic events triggered the Uralic spread’, Diachronica , vol. 39, no. 4, pp. 490–524, Aug. 2022, doi: 10.1075/dia.20038.gru. [35] A. Childebayeva et al. , ‘Bronze age Northern Eurasian genetics in the context of development of metallurgy and Siberian ancestry’, Commun Biol , vol. 7, no. 1, p. 723, June 2024, doi: 10.1038/s42003-024-06343-x. [36] S. Grunwald, ‘Riskante Zwischenschritte: Archäologische Kartographie in Deutschland zwischen 1870 und 1900’, in Massendinghaltung , B. Hofmann and others, Eds, Leiden: Sidestone Press, 2016, pp. 111–147. [37] E. Herrera Malatesta and S. De Valeriola, ‘Ambiguous landscapes: A framework for assessing robustness and uncertainties in archaeological point pattern analysis’, PLoS ONE , vol. 19, no. 9, p. e0307743, 2024, doi: 10.1371/journal.pone.0307743. [38] H. Müller-Karpe, Die Urnenfelderkultur im Hanauer Land . Marburg: Elwert-Gräfe und Unzer, 1948. 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 [39] Hessisches Landesamt für Bodenmanagement und Geoinformation, ‘Verwaltungseinheiten Hessen. Geobasisdaten’. 2025. Accessed: Apr. 03, 2025. [Online]. Available: https://gdk.gdi-de.org/geonetwork/srv/api/records/e72c607e-304d-49db-afa1-06f2 49a4466a [40] R Core Team, R: A Language and Environment for Statistical Computing . Vienna, Austria: R Foundation for Statistical Computing, 2025. [Online]. Available: https://www.R-project.org/ [41] A. Baddeley, E. Rubak, and R. Turner, Spatial Point Patterns: Methodology and Applications with R . Boca Raton: Chapman and Hall/CRC Press, 2015. [42] S. Garnier, N. Ross, B. Rudis, A. P. Camargo, M. Sciaini, and C. Scherer, viridis: Colorblind-Friendly Color Maps for R . 2021. [Online]. Available: https://CRAN.R-project.org/package=viridis [43] R. J. Hijmans, terra: Spatial Data Analysis . 2025. [Online]. Available: https://CRAN.R-project.org/package=terra [44] E. Pebesma, ‘Simple Features for R: Standardized Support for Spatial Vector Data’, The R Journal , vol. 10, no. 1, pp. 439–446, 2018, doi: 10.32614/RJ-2018-009. [45] T. L. Pedersen, patchwork: The Composer of Plots . 2020. [Online]. Available: https://CRAN.R-project.org/package=patchwork [46] H. Wickham, ggplot2: Elegant Graphics for Data Analysis . New York: Springer-Verlag, 2016. [47] H. Wickham, R. François, L. Henry, and K. Müller, dplyr: A Grammar of Data Manipulation . 2023. [Online]. Available: https://CRAN.R-project.org/package=dplyr [48] H. Wickham and M. Girlich, tidyr: Tidy Messy Data . 2023. [Online]. Available: https://CRAN.R-project.org/package=tidyr [49] F. Thiery and F. Schenk, ‘Campanian Ignimbrite Geo Locations’, Squirrel Papers , vol. 5, no. 2, p. #2, 2023, doi: 10.5281/zenodo.10361309. [50] F. Thiery, ‘Fuzzy Spatial Locations Ontology’, Squirrel Papers , vol. 5, no. 2, p. #3, Dec. 2023, doi: 10.5281/zenodo.10362777. 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689