CIDER-CL A system for monolingual and cross-lingual ontology alignment

CIDER-CL (Cross-lingual CIDER) [1] grounds on CIDER (Context and Inference baseD ontology alignER). In its current version, cross-lingual capabilities have been added by including the use of Cross-Lingual Explicit Semantic Analysis (CL-ESA) [2] between entities documented in different natural languages. For monolingual comparison we use SoftTFIDF measure [3]. Artificial neural networks are used to combine elementary similarity computations.

CIDER-CL can operate in two modes:

• Ontology aligner: taking two OWL ontologies as input and giving their alignment as output in RDF.
• Similarity service: taking two ontology entities as input and giving the similarity value between them as output.

The type of alignment that CIDER-CL obtains is semantic equivalence. In its current implementation the following languages are covered: English (EN), Spanish (ES), German (DE), and Dutch (NL). CIDER-CL is not intedeed to be used with large ontologies, particularly in the cross-lingual case.

The ontology alignment process carried out by CIDER-CL is as follows:

1. CIDER-CL receives two ontologies as input (O1, O2 in the figure). The ontological context of each ontology entity is extracted and enriched by applying a lightweight inference mechanism.
2. Similarities are computed between different features of the ontological context (labels, comments, subterms, superterms, properties, etc.). SoftTFIDF and CL-ESA measures are used for monolingual and cross-lingual comparisons respectively.
3. The different similarities are combined within artificial neural networks to provide a final similarity degree for each pair of ontology entities.
4. A matrix (M) with all similarities is obtained. The fnal alignment (A) is then extracted from this matrix, fnding the highest rated one-to-one relationships above a certain threshold.

Prototypes and demos:	CIDER-CL is available for download on Github . The CL-ESA implementation used by CIDER-CL is also available here
Experimental evaluation:	CIDER-CL participated in the OAEI'13 evaluation campaign
Related publications:	Click here
Contact:	Jorge Gracia at

References

[1] J. Gracia, K. Asooja, "Monolingual and Cross-lingual Ontology Matching with CIDER-CL: Evaluation Report for OAEI 2013", Proc. of OM’11 workshop @ISWC’13.

[2] P. Sorg and P. Cimiano. “Exploiting wikipedia for cross-lingual and multilingual information retrieval”. Data Knowl. Eng., 74:26–45, Apr. 2012

[3] W.W. Cohen, P. Ravikumar, and S. E. Fienberg. “A comparison of string distance metrics for name-matching tasks”. In Proc. of IIWeb-03 workshop @ IJCAI’03