Modelling Knowledge for Independent Performing Arts: a Semantic Wiki in Intangible Cultural Heritage Case Study

Tracking #: 2362-3575

Diego Torres
Leandro Antonelli
Jonathan Martin
Alejandro Fernandez
Mario Lezoche
Mariana del Mármol

Responsible editor: 
Special Issue Cultural Heritage 2019

Submission type: 
Full Paper
Performing arts as Dance, Theater, and Music are considered by UNESCO as part of the intangible cultural heritage. There is a branch called independent performing arts that refers to performing arts outside of the commercial and official circuit. Independent performing arts are characterized by being produced with reduced infrastructure and economic resources. This situation foster the "creative freedom" that give origin to tremendous artistic diversity and multiplicity of performers and plays. This article tackles the challenge of studying and recording the independent performing arts in the city of La Plata in Argentina. This study focuses on a semantic infrastructure to collect and describe theater, dance, and performances in the independent circuit. This work introduces a collaborative platform called Nodos that stimulates collective knowledge building and sharing. The platform uses semantic web technologies to capture the domain enabling computer-based processing. Thus, it makes it possible to derive new information through the inference procedures.
Full PDF Version: 


Solicited Reviews:
Click to Expand/Collapse
Review #1
By Mikko Koho submitted on 13/Feb/2020
Review Comment:

The article describes a catalog of the independent performing arts in the city of La Plata. Along with the catalog, the article presents an OWL ontology called NodosOntology, which is used as the metadata schema for crowdsourcing information using Semantic MediaWiki (SMW). The article presents statistics of the resulting data and social network analysis of the people described in the wiki.

The topic is relevant to the Semantic Web for Cultural Heritage special issue. Cultural heritage is in the focus of the study and Semantic Web technologies are employed in creating an ontology to represent crowdsourced information. The idea of representing intangible cultural heritage as Linked Data is timely and relevant. However, the study described in the article could have been conducted just as well without the use of SW technologies. The article might be good fit for a more general Digital Humanities publication.

The article doesn't contain explicit research questions, but a general notion of the research problem of "tackling the challenge of studying and recording the independent performing arts in the city of La Plata in Argentina". The research contributions are a description of an ontology, the use of a system for recording and studying independent performing arts, and evaluation of the collected information. There is not much originality in the ontology development or use of Semantic MediaWiki to collect crowdsourced information but some originality in the social network analysis of the information in this domain. The modeling of information about intangible cultural heritage has been discussed previously in more depth than considered by this article.

The lessons learned from this project could be useful for other projects that are crowdsourcing cultural heritage information. The system has been running since 2016, but the practical significance and adoption of the system remains unclear. The article is written in a clear and understandable way and the text is organized logically. The paper would, however, benefit from further proof-reading.

Major remarks

The ontology development is said to follow [9] (Noy & McGuinness: Ontology Development 101), which provides a good basis for ontology development. However, the NodosOntology creation is previously reported in a journal article about the Nodos project, [7], where the creation is explained to follow a different ontology development procedure. The relationship between the ontology versions is unclear, but the changes appear to be minor.

On a general level, the presented NodosOntology metadata schema is efficient for the purposes described. It is a simple data model for annotating basic information about, e.g., people, plays, and events. There could be more discussion and even use of commonly used standards in the CH sector, such as, CIDOC CRM, FRBR, or FRBRoo. The classes are linked mostly to Luke Blaney's Theatre Ontology, which is a part of the Linked Open Vocabularies. The properties do not seem to be linked to any existing vocabularies. The classes, properties, and their use should be explained in more detail, now it remains unclear what are some classes depicting, e.g., Season, FestivalPresentation, SpecialShow, and Edition. Are there any relationships between the classes?

The equivalence of the Play class with the Luke Blaney's Theatre Ontology's Production class, and its implications, should be discussed. The Production class is a sub class of FRBR Expression class, rooting a core class of the NodosOntology to the widely used FRBR standard. Representing live performances with FRBR is not very straight-forward as discussed in (David Miller & Patrick Le Boeuf (2005) “Such Stuff as Dreams Are Made On”: How Does FRBR Fit Performing Arts?) and (Doty, C. (2013) The difficulty of an ontology of live performance). NodosOntology's Play class seems to refer to productions of a play, which according to the aforementioned publications would correspond more to the FRBR's Manifestation class, whereas the script of a play could be seen as the Expression. This seems like a fundamental problem with the conceptual modeling of the domain in NodosOntology, but there is not enough information in the paper to understand the created conceptualization.

Additionally, the core class Event is defined to be equivalent with the Luke Blaney Theatre Ontology's Event class. However, the Theatre Ontology does not define an Event class. It contains a reference to Event class, which looks like an error both in the Theatre Ontology and in NodosOntology. The used Theatre Ontology URI namespace is not correct and it should be as given in the LOV web page. There is no explanation what the core class Event is used for. According to Table 2 it has no data properties.

On page 6, the whole RDF data dump of the Nodos wiki is confusingly referred to as NodosOntology. The data dump does not seem to contain the ontology described in Chapter 3, but in addition to the Nodos data, seems to contain plenty of spam, such as the description of the wiki page . Is the ontology available somewhere for human and machine users? The paper should contain the URI of the ontology description and URIs of the described resources.

The resulting dataset is evaluated using statistics and standard social network analysis methods, providing some evidence that the data itself is useful as an artifact. The statistics of individuals by class contain some amount of empty pages that are linked to in the wiki. Is it possible to filter the empty pages out to get more sensible statistics? The statistics relating to properties contain mostly SMW internal definitions instead of NodosOntology properties. It remains unclear what part of the statistics are not SMW internal definitions. It is not possible to understand the statistics shown in Table 5 without further explanation of the categories.

The social network analysis results are compared against conducted ethnographic studies for support and contextualization, with results that sound promising. However, some discussion concerning the validity of the network analysis results should be presented. How well does the data actually cover the domain? Are there biases in the data? Has there been common practices of adding informain to the wiki? Does the data contain all the relevant connections between the contained people and their groups and plays?

There should be more related work about Digital Humanities projects collecting information via online platforms such as SMW, and additionally about studies modeling and publishing Linked Data about performing arts.

Some minor remarks

- Text in the Figures is mostly way too small
- In "3.1 Domain and scope in a Learning and Design process" I expected to read about the ontology domain and scope as the ontology development procedure would suggest. However, the ontology domain is only vaguely discussed and the ontology scope is not discussed at all. What does the learning and design process mean?
- The abbreviation "GEC" is used on page 8 without explanation of what it means
- There is variation in how references are presented

Review #2
Anonymous submitted on 19/Mar/2020
Review Comment:

The paper presents the NodosOntology and its use case for collecting and representing a performing arts catalog in Argentina, and an analysis of its content. The ontology is used in a semantic MediaWiki application called Nodos for recording arts performances in a festival series, not really for modeling the contents of the performances, such as events.

First the general idea of storing data about art performances and the use case are presented (sections 1-2). Then a fairy informal account of the NodosOntology is given (section 3). From a theoretical and methodological point of view, this ontology does not bring in new contributions and the ontology is not rigorously represented.

Semantic Media wiki is used for annotating contents using the ontology in the Nodos application by crowdsourcing. This is nice but there is nothing particularly novel in this approach. The data is not enriched with linking to other data services but the ontology reuses URIs from some well-known vocabularies, such as FOAF.

In section 6, statistics of the data aggregated in the Nodos system are presented giving some insights into the data and its evolution over several years of usage. It is nice to learn that an application like this is really in use. There is also a subsection showing some interesting network analytics of the persons present in the data, a nice benefit of having the crowdsourced Wikidata available.

This manuscript was submitted as 'full paper' and should be reviewed along the usual dimensions for research contributions which include (1) originality, (2) significance of the results, and (3) quality of writing.

As for originality, I do not see much originality in the ontology design. The use case and idea of analyzing the data is more original, but is this is enough for a scientific paper in the SW journal remains questionable.

The results are significant in the sense that the solution presented seems to work and is actually in use in the use case presented. However, from a more general methodological and theoretical viewpoint, the paper does not provide much scientific contributions even if it presents a nice application case.

The writing quality is acceptable, but careful proofreading is needed for typos here and there.

In my mind, the paper and the case study are interesting but based on existing methodology and technical solutions. The problem of the paper in the SWJ context is lack of scientific novelty and contributions as explained above. Therefore, I think SWJ is not the place for publishing the results. I would suggest submitting the paper to , e.g., a more culturally oriented journal where the point of using semantic web could already be of interest.

Review #3
By Beatrice Markhoff submitted on 13/Jul/2020
Review Comment:

This manuscript was submitted as 'full paper' and should be reviewed along the usual dimensions for research contributions which include (1) originality, (2) significance of the results, and (3) quality of writing.

This article presents a platform called Nodos for a collaboratively built catalog of independent performing arts in La Plata, Argentina. This platform is a semantic mediawiki, based on a simple ontology comprising 9 classes, mainly Plays, People,
Groups, and Venues, and few properties, e.g. beginning and end dates, participants to plays and groups. The instances are created by users when they collaboratively write articles on plays, groups and performers, via web forms as usual in semantic mediawikis. In Introduction the authors write that this ontology is to support semantic reasonning but this is not illustrated in the paper. Wikis are powerfull when they are actually used by a community of contributors: this is not demonstrated when visiting the platform. One interesting point of the article is an analysis of the actors relationships, but it is something that can be done without using ontology. Regarding the quality of writing, it is rather correct but requires proofreading.

To sum up, for readers who are interested in Semantic Web for Cultural Heritage this article is too limited. More precisely, there is no research contribution for the semantic web community, as the article does not bring anything new compared to [1] for instance, and, for the cultural heritage community, the benefits of the semantics in the presented solution are not clearly demonstrated.

[1] Schaffert, S. (2006). Ikewiki : A semantic wiki for collaborative knowledge management. In15th IEEE International Workshops on Enabling Technologies: Infrastructures for Collaborative Enterprises (WETICE 2006), 26-28 June 2006, Manchester, United Kingdom, pp.388–396. IEEE Computer Society.