The POSTDATA Network of Ontologies for European Poetry

Tracking #: 2428-3642

Salvador Ros
Maria Luisa Diez Platas
Elena González-Blanco
Oscar Corcho
Helena Bermúdez
Javier de la Rosa
Alvaro Pérez

Responsible editor: 
Special Issue Cultural Heritage 2019

Submission type: 
Ontology Description
A relevant line of work in Digital Humanities is related to standardization processes to describe traditional concepts using computer-readable languages. In regard to Literary studies, poetry is a particularly complex domain due to, among other aspects, its special use of language. This paper presents a network of ontologies for capturing domain knowledge about poetry, including the poetic work itself, as well as its structural and prosodic components. The network also includes an ontology about dates, focusing on the special representational needs of literary works. This work is one of the results of the POSTDATA (Poetry Standardization and Linked Open Data) European Research Council project, which aims to provide a means for researchers on European poetry to publish and consume semantically-enriched data.
Full PDF Version: 

Reject (Two Strikes)

Solicited Reviews:
Click to Expand/Collapse
Review #1
By Stefano Borgo submitted on 18/Mar/2020
Review Comment:

The paper has improved considerably. It's starting point and aims are much clearer now. The presentation of the ontology is fair.

A few minor issues:

pg.2 sentence to fix (although-however; working [on])
"Although the relationship between onto- logical models and TEI has been taken into consideration in recent years [22–24], however, it is an open field that we are still working because the description of poetry in TEI has to improve yet."

Fig. 2 what is the meaning of the three arrows without label?

pg.9 "ibjetc"?

pg.13 Annex II
"A characteristic that should be highlighted is that this poetic composition is a song, with a refrain. The refrain is the first stanza, but it is repeated after all the others. For this reason, the refrain has no number assigned to it and its lines are indicated as part of a stanza."

This text repeated twice in the page with inconsistencies (e.g. compare: "For this reason, the refrain has no number assigned to it and its lines are indicated as part of a refrain.")

Review #2
Anonymous submitted on 28/Apr/2020
Major Revision
Review Comment:

I did not change my previous recommendation because the new version of the manuscript is just a slight update of the original submission. The authors took into account the big majority of my remarks (as well as the ones of the other reviewers) citing other works or deleting the problematic sentences (that indeed highlighted some problematic and unmotivated choices) instead of adding some elucidations and justifications of the undertaken choices. This move slightly improves the presentation but the paper remains incomplete and imprecise.

Actually, the whole role of the paper is not clear. In the reply to the reviewers the authors claim "The methodology that we followed for the development of the ontology is described in the references below. In this paper, we preferred focusing on how the ontology network is structured so that it can be easily used by somebody who wants to generate data according to this ontology network or make use of data already available according to this network." This contrasts with the sentence present in the paper "This paper presents the methodology carried out to build such a network of ontologies and the most significant ontologies of this network" as well as with the fact that sections 2 and 3 are not devoted to the presentation of the ontology network. In any case, I don't see how this paper might allow a reader to "easily" and correctly use the proposed ontology network. The description of the classes, relations, and constraints in the modules is too partial and imprecise without any attempt to deep the conceptual/ontological analysis. In addition, the few examples considered are modeled in way that, to me, is unclear and ad hoc (see my comments below). This fails to satisfy the requirement for papers of kind "description of ontologies": "Illustration, clarity and readability of the describing paper, which shall convey to the reader the key aspects of the described ontology." Concerning the requirement about "Quality and relevance of the described ontology", the comparison with other works in this field is partial and it does not clearly point out what are the improvements, and the evaluation of the proposed ontology in terms of quality parameters and of its role in applications or in use-case experiments is still very partial (as admitted by the authors in the reply of the first reviews).

Concerning the examples, my previous question about the example in fig.5 has not been answered (and Fig.5 did not change). I copy below my original question:

"I'm not sure to understand the example depicted in fig.5. Is this an example of a stanza composed by two lines: "Mo chion dot..." is the (content of the) first line and "a mhic na flatha..." is the (content of the) second line? If this is the case, first I don't see why "a mhic na flatha..." is attached to the stanza [11161] and not to the line [11152]; second I don't see why only the first line is attached to OrderedLineList[1115] or, vice versa, why not attaching both lines to OrderedStanzaList[1116]."

I think this is quite confusing and it does not help in clarifying the modeling options offered by the proposed ontology. I suggested to clarify and illustrate the ontological notions in the modules by means of a unique and complete running example but the authors decided to leave the example in fig.5 largely underspecified and to add a different example in the appendix. I still think that a running example informally introduced at the beginning of the paper and represented, step by step, using the different modules in the ontology network would be much more effective.

The example introduced in the appendix concerns the modeling of a song with a refrain. The solution adopted for representing the refrain sounds to me ad hoc and not easily generalizable.
The OrderedStanzaList seems not expressive enough to represent lists where the same Stanza can appear several times in different positions. For instance, one can imagine a poem or a song with the following stanza structure: 1 2 3 1 3 1 2 1 1. Here there is not a refrain and there is not a clear repeated pattern of stanzas, still the same Stanza can appear in different "positions" in the StanzaList. It is then not clear why the authors decided to reuse the module of ordered lists rather than extend or modify it to allow to associate to the same Stanza e set of positions (an not a unique one). And this is just one specific conceptual point one could at least discuss.

Review #3
Anonymous submitted on 07/Jun/2020
Major Revision
Review Comment:

As far as I was able to establish most of my major misgivings about the paper were not properly addressed by the authors. On the other hand, most of minor comments, related to to typos and grammar, were adequately met.

As for the need to use a top-level ontology I cannot see how the authors' rebuttal addresses my comment. My point here was not about the clarity or completeness of the paper's exposition or about the way other ontologies's categories were brought into postdata ontologies.

I think the authors should include the discussion of the multiple parents for the domain specific categories even if it is, as they say, a common practice in the community.

Contrary to what the authors say in the cover letter the problem with ExactDateExpression has not been fixed, i.e., this class is still not satisfiable. At the time of my review I couldn't download the source of the core ontology from - got a 404 error when I tried to download the source for each format mentioned in this page.
However, postdata-dates is still available from And the latter ontology has the aforementioned problem. Because of this problem with the access to, I was unable to check whether other implementation problems have been fixed.

I appreciate that the authors provided a running example, although the way they defined the URIs for the individuals makes it difficult to follow/read: multiple individuals in the example have the local part equal to Cantiga3-1, i.e., they end with the same string 'Cantiga3-1'. Also I am not sure what the recurring comment "# The remaining lines are omitted due to code extension" is supposed to mean.