Review Comment:
The manuscript presents BISTÌRIS, an ontology built to model Sardinian traditional costumes as a form of intangible cultural heritage. The paper details the methodology used to develop the ontology, including its conceptual structure, the integration of components from other existing ontologies, population into a knowledge graph (KG), and evaluation via reasoning-based validation and competency questions. The work situates itself within the broader field of semantic web applications for cultural heritage preservation and digital humanities. The manuscript clearly presents the ontology and its specific cultural focus on traditional costumes, addressing a previously identified gap in this field. The manuscript argues for the necessity of a domain-specific ontology for traditional costumes, as opposed to more general cultural heritage (CH) ontologies (e.g., CIDOC CRM) or existing efforts dedicated to clothing (e.g., ArCo Clothing Description Ontology, Costume Core metadata schema), which appear limited for adequately describe the complexities of Sardinian costumes in particular. The manuscript provides a detailed, structured ontology with clear class definitions and a hierarchy that enhance transparency and reproducibility. It also demonstrates the population into a functioning Knowledge Graph (KG) with queries, publicly available via a dedicated SPARQL endpoint. While the manuscript is well-structured and clear, minor adjustments could further strengthen it.
The ontology is innovative in its specific cultural focus and has interesting potential applications. Although in the introduction the manuscript claims adaptability to other traditional costumes from different Mediterranean traditions (“Although rooted in the Sardinian context, the presented ontology is conceived to be adaptable for describing other traditional dress systems in the Mediterranean area – such as those found in Greece or Southern Italy – thus offering a common framework for comparative studies and cross-regional heritage initiatives”), no demonstration or case study beyond Sardinia is included. The authors could clarify how this adaptability is conceived, taking into consideration both the variability of the attire elements in other traditions, as well as different languages (e.g. Greek). Related to the language, several terms are in Italian. The authors state their decision to keep certain terms in their original language, rather than translating into English, to preserve the precision and clarity. Maybe the authors could consider including a glossary. For instance, if this reviewer is correct, there is no description for “berritta”.
Moreover, with regard to the ontology, it would be interesting to know if there are other elements adorning the traditional dresses, such as specific buttons or pins to hold the fabric together. Do such elements exist in Sardinian traditional dresses? If so, they are not included in the ontology classes and it would be interesting to include them in the model as part of the attire, to possibly understand through KG if they have different meanings or purposes. This reviewer was also wondering if, beyond the male and female categories, there were different garments for people of different ages or connected to specific community roles. If so, this could be another element to express through the data properties.
Concerning the integration of elements from other existing ontologies in the domain, the authors might consider aligning the proposed ontology with a more general CH ontology, such as CIDOC CRM (see for instance, Catalano et al. 2020 - https://doi.org/10.1007/s00799-020-00287-3).
This reviewer has identified some further points in the different sections of the manuscript and added comments. In the Introduction, the sentence “Among the many […] through visual and material expression”, could be accompanied by references to general anthropological studies on traditional costumes. Regarding the sentence “Digital libraries such as Europeana now host a wide range of items related to traditional costumes, including those from Sardinia.”, it would be more helpful to include a link to the Sardinian traditional costume items on Europeana rather than a general link to the Europeana website. The authors mentioned that Europeana aggregates items on traditional costumes. How are these items described in their original databases, and what metadata is used? Are there databases in Europe dedicated to archiving traditional costumes? If so, what metadata or ontology is used? In the sentence “under the activities of Spoke 2”, please add a reference or link, for readers who are unfamiliar with it. In the sentence “A notable example is ArCo, the Italian Cultural Heritage Knowledge Graph […] items from the national catalogue”, to which national catalogue are you referring to? Please specify and add a reference/link.
The discussion in the section dedicated to related works could be improved by distinguishing more clearly between ontologies, metadata schemas and knowledge graphs.
In Section 3.3 (p. 8), why do some of the garments not belong to F-Garment in the list of classes? The same applies to M-Garment.
In Section 4.1, under Q1, the reviewer was confused about the source of this information: it may be helpful to refer to Section 3.5.3, 'Data Properties', where you mention that geographical data provides details about the geographical origin of the exhibited customs, including longitude, latitude and height above sea level.
Following, there are some more general remarks. Check the punctuaction (it is missing in a couple of places). For CIDOC CRM, the correct spelling is without the hyphen. Please check the acronyms. For example, the term 'KG' is recognisable, but it is never explicitly stated in the text that 'KG' stands for 'Knowledge Graph'. The same applies to Semantic Web Rule Language (SWRL), Galleries, Libraries, Archives, and Museums (GLAM), and competency questions (CQ). In English, it is 'dating' or 'date', not 'datation'.
|
