Survey on complex ontology matching

Tracking #: 2164-3377

Elodie Thieblin
Ollivier Haemmerlé
Nathalie Hernandez
Cassia Trojahn dos Santos

Responsible editor: 
Marta Sabou

Submission type: 
Survey Article
Simple ontology alignments, largely studied in the literature, link a single entity of a source ontology to a single entity of a target ontology. A limitation of these alignments is their lack of expressiveness which can be overcome by complex alignments. While diverse state-of-the-art surveys mainly review the matching approaches in general, to the best of our knowledge, there is no study of the specificities of the complex matching problem. In this paper, a review of the different complex matching approaches is provided. It proposes a classification of the complex matching approaches based on their specificities (i.e., type of correspondences, guiding structure). The evaluation aspects and the limitations of these approaches are also discussed. Insights for future work in the field are provided.
Full PDF Version: 


Solicited Reviews:
Click to Expand/Collapse
Review #1
By Antoine Zimmermann submitted on 04/Apr/2019
Review Comment:

The new revision addresses my concerns satisfactorily. I appreciate that the authors identified clearly what was modified, which helped me a lot in evaluating this version. I only have the following small remarks that should be solved easily for the final version:

Minor comments:
- Definition of correspondence: either a correspondence is a triple, and there are 3 components, or it is a quadruple, and there are 4 components, or it can be a triple or a quadruple. It cannot be a triple that can have a fourth element. Considering that examples of correspondences in the paper are now always triples, I would suggest to mention the confidence value outside the definition of a correspondence (e.g., "alignment systems usually assign a confidence value to each correspondence, such that correspondences are sometimes defined as quadruples (e1,e2,r,c). We only exemplify correspondences as triples in the rest of the paper.") This way, I think it addresses Reviewer 1's comment on this and stay true to what's actually used in the examples.
- in Sec.6.3: "pattern-based alignment formats such as EDOAL (or OWL)" -> OWL is not an alignment format.
- Section 7: "type of correspondence" -> correspondences

Review #2
Anonymous submitted on 22/Apr/2019
Review Comment:

I would like to thank the authors for the new version. Though I had said I was already trustful in the last review, that they could fix the remaining problems, I am giving further feedback now.

I acknowledge that the authors have tackled the minor comments/suggestions I had made in my earlier review. It seems that the paper has greatly improved.

There are still aspects that I am not very satisfied with, for example I still think section 2.4 on scope clarification is unnecessary long, and I don't fully get the gist of the various visualisation efforts in 3.2. But it is probably up to the author to decide how to treat their reader now. I would however argue that another proofread is welcome, as it seems the one that has been done between version 2 and 3 has been done before some additions in 3. Therefore, some editorial issues re-appear, such as in definition 1 which has two main verbs ("is"), or a sentence on p31 that just doesn't make sense to me (the one with "regarded as perspective")

Some other comments on new parts:

On p6 I don't see how the SPARQL SELECT queries may qualify as a mapping. They just extract resources from data patterns in the target and destination datasets, but don't express (or process) any correspondence (as opposed to the following CONSTRUCT example)

Table 1 could be ordered to follow the order in 3.1, and I don't understand why the xR2RML line has "mixed formats" while the D2RML has not (it seems like it's the same kind of format enumeration)

Review #3
By Catia Pesquita submitted on 06/May/2019
Review Comment:

The authors have successfully addressed my previous concerns.