The Bowlogna Ontology: Fostering Open Curricula and Agile Knowledge Bases for Europe’s Higher Education Landscape

Paper Title: 
The Bowlogna Ontology: Fostering Open Curricula and Agile Knowledge Bases for Europe’s Higher Education Landscape
Gianluca Demartini, Iliya Enchev, Joël Gapany and Philippe Cudré-Mauroux
The Bologna Process initiated a radical change within higher education institutions. This change triggered the creation of new administrative procedures in the every day life of European universities. It also gave rise to the emergence of new concepts for the description of curricula. It is critical for the successful continuation of this process to support the publication and exchange of information among universities. With this aim in mind, we created the Bowlogna Ontology to model an academic setting as proposed by the Bologna reform. In this paper, we present our efforts to design this ontology and the entire process that lead to its creation starting from the definition of a linguistic lexicon derived from the Bologna reform and its conversion to a formal ontology. We also describe practical applications of our ontology for end-users at universities (such as a faceted search and browsing system for course information).
Full PDF Version: 
Submission type: 
Ontology Description
Responsible editor: 
Guest Editors

This is a revised manuscript, submitted as an Ontology Description paper. The original submission received a reject and resubmit decision and had been submitted as a full paper. Below the reviews of the revision, followed by those of the original submission.

Solicited Review by Carsten Keßler:

This revision addresses my comments for the initial submission. The only thing that would add to the quality of the contents would be an explanation of the colors of the arrows in figure 1, as it is not quite clear to the reader for which kinds of relations the different colors stand.

Some typos to fix for the final manuscript:
- Introduction, 3rd line: declaration,. -> remove comma
- sec 4.2, missing point after "Protege editor" (should also be Protégé, by the way)
- p. 10, top left column: "transitioning form" -> from

Solicited Review by Mathieu d'Aquin:

This paper presents the development of the Bowlogna ontology, to support the interlinking and interoperability between Universities, based on the Bologna process.

First I think that this is a very interesting development, that could end-up being really useful to the community. I particularly appreciate that the authors have clearly described the methodology used to build the ontology.

The small discussion on related ontologies is also very useful, and could help in the adoption of the ontology. I would encourage the authors to provide, alongside the ontology, mappings to some of these existing vocabularies (especially MLO).

The main "validation" element of the paper regarding the ontology (as far as I can see) are the presented use-cases. I understand that, in principle, not much more can be done at this stage to show the value of the ontology. It is however slightly disappointing that this is really only showing that an ontology is needed, not that this one in particular is specially good. Also, in my view, only the third one is really related to the main motivation for the ontology. Having said that, I think that this is a very important use case, and that relying on Bologna to realise it is actually a good and valuable approach.

I would have liked to see more of the value of the ontology demonstrated through actual usage: has it been employed by some universities? Are their plans for this adoption?
I think these are important questions: the ontology might be super good, but if it is not used, it does not really have much value. This is mentioned in the "future direction" section, but I would really want to see a more substantiated discussion here, about concrete plans to spread the use of the Bowlogna ontology (as I would be myself be interested in using it concretely).

Solicited Review by Anonymous Reviewer:

The revised version of the paper addresses some of the issues raised by the referees (such as removing the already published material about the benchmarking etc.) and now seems to be more focused on the ontology creation process.

It is good that the authors added more discussions about public/private parts of the ontology and multi-linguality, but it is also remarkable that the discussions are not really in-depth. Similarly, the revised paper includes a section about related ontologies, but the motivation for designing a new ontology as proposed in this paper still remains unclear.

Reviews of the original submission:

Solicited Review by Carsten Keßler:

The paper introduces the "Bowlonga" ontology, an OWL ontology to describe higher education environments based on the Bologna process. The paper is well written and clearly describes the ontology, the design process and rationales, and presents use cases and an evaluation method. I recommend to accept the paper if the authors address the following issues:

1. If possible, it would be nice to have the lexicon developed in the process available online. This would help a lot to follow and judge the design decisions made when creating the ontology from the lexicon.

2. In section 4.3, it would be nice to have some more examples that show how the different mereology relations were used in the ontology.

3. Section 5: "Based on our understanding of the Bowlogna reform…" – has there been any kind of evaluation whether this understanding is shared by others, i.e., whether other experts familiar with the Bologna process agree on your modeling in the ontology? This would be particularly interesting for the multi-lingual aspects; in general, the multi-linguality of the ontology is a bit underrepresented in the paper, some more details would certainly be of interest for the reader.

4. I like the idea if private and public ontology parts, as it is clearly necessary in this case. However, some more details about how this was implemented would be helpful: I assume that the ontology only indicates which parts are private and public, resp., and it is left up to applications to make sure that private parts are not exposed publicly?

5. The section on the Bowlogna Benchmark would benefit a lot from a sample evaluation; with the oure description of the benchmark, as it is right now, it is not very useful. I would recommend to either add a sample evaluation or consider to remove this section, and list the sample queries somewhere else.

6. Some references to ontology engineering methods such as OntoClean would be nice – did you use any of them, and if not, why?

7. Typos/language:
- p.1, right column "they credits the earned" -> "the credits they earned"
- p.4, left column "step step"
- p.10, right column: "can be leverage to foster information exchange" -> this should probably be "leveraged", but even then, the sentence reads a bit strange. Consider revising.

Solicited Review by Mathieu d'Aquin:

This article presents an initiative towards building an ontology to represent curricula in european higher education institutions, that would be based on the bologna declaration. This is an interesting area and I believe that this ontology could be very useful in many applications, and to enable the exchange of information across institutions. The paper would however need to address major issues that have not been considered in the current version in order to be acceptable.

One of the main comment I would have is that it is very unclear from the article what the ontology contains and what it can be used for. There is a link to the ontology, but no documentation is attached to it. The authors say that "we can imagine different usage scenarios" and describe a few of them in section 7, but these scenarios are not specific to the ontology (faceted browsing can be done with any ontology) or the domain. Also, most of what is described already exists in universities currently. The added value brought by the ontology should be the focus, and is currently very unclear. More importantly, the use of the ontology for benchmarking is being described, but that obviously does not say anything about the quality and usefulness of the ontology, in relation with the education domain. I would have expected from a paper about an ontology to have more description and justification of the specific modelling choices that were made.

While the methodology to create the ontology is described and interesting (even though I would have liked to know a bit more about the results of each step), the motivation for the authors to create this ontology remains unclear. Also, there are many other ontologies (/schemas/vocabularies) that could have been either reused or mapped to (e.g., LMO, XCRI, DC, AIISO, courseware) and that are not even mentioned in the article.

Solicited Review by Anonymous Reviewer:

This is an interesting work that describes the steps of creating a lexicon and then an ontology based on Bologna framework concepts. Next, this ontology is employed to create an RDF benchmark (for research usage) and in a prototype faceted-search system (for end-users at universities).

In its current stage, the level of contribution seems to be inadequate for a journal publication. First of all, a considerable part of this work (describing the creation of ontology and RDF benchmark) is already published as a symposium paper, as properly cited by the authors (i.e., reference [6]). The new material, presented in Section 7, only describes some use cases for the developed ontology. Furthermore, I am not totally convinced by the problems described in Section 7.2. Given that many large-scale institutions use traditional database systems for similar purposes, the problems mentioned in this section seem to be caused by somewhat inappropriate design choices, rather than being inherent problems of the systems. (This, of course, does not mean that linked data approaches can not improve such traditional systems, but I only think that the mentioned problems do not necessarily represent the points that could require improvement in this setup.)

From a technical point of view, the application domain is rather restricted (i.e., the resulting ontology includes only 66 classes). Other than the ontology creation process, which is essentially carried out as a manual task, there seems to be no novel contributions in terms of the methodology, algorithms, evaluation, etc. I think that the paper should be developed in these directions to justify as a journal publication.