BLOOMS

From Knoesis wiki
Revision as of 03:24, 2 August 2010 by Prateek (Talk | contribs) (Alignment API Linked Open Data schema Alignment Performance)

Jump to: navigation, search

BLOOMS is an ontology alignment system based on the idea of bootstrapping information already present on the LOD cloud. It was developed particularly for Linked Open Data schema alignment.

To obtain more information about BLOOMS, please have a look at our paper Ontology Alignment for Linked Open Data.

BLOOMS is an acronym for Bootstrapping-based Linked Open Data Ontology Matching System.

Approach

BLOOMS bootstrapping approach utilizes the Wikipedia category hierarchy for aligning ontologies. BLOOMS constructs a forest (i.e., a set of trees) TC (known as BLOOMS forest for C) for each matching candidate class name C, which roughly corresponds to a selection of supercategories of the class name. Comparison of the forests TC and TB for matching candidate classes C and B then yields a decision whether or not (and with which of the candidate relations) C and B should be aligned.

Evaluation

We performed a comprehensive evaluation of BLOOMS using third party datasets and other state-of-the-art systems in ontology matching. More specifically, BLOOMS has been evaluated in two different ways.

  • We examined the ability of BLOOMS to serve as a general purpose ontology matching system, by comparing it with other systems on the Ontology Alignment Evaluation Initiative (OAEI) benchmarks.
  • Secondly, we evaluated BLOOMS for the purpose of LOD schema integration and compared it with other systems for ontology matching on LOD schema alignment.

For both the evaluations BLOOMS has been compared with the state of the art tools in ontology mapping.

Systems for Comparison

  • RiMOM: RiMOM was the top system in the oriented track of OAEI in terms of f-measure and availability for download.
  • AROMA: AROMA ranked second in the 2008 OAEI Benchmark event.
  • OMViaUO <ref name=omviauo-2009>Mascardi, V., Locoro, A., and Rosso, P. 2010. Automatic Ontology Matching via Upper Ontologies: A Systematic Evaluation. IEEE Trans. on Knowl. and Data Eng. 22, 5 (May. 2010), 609-623. DOI= http://dx.doi.org/10.1109/TKDE.2009.154</ref>: OMViaUO utilizes upper level ontologies such as SUMO and DOLCE as semantic bridges in the ontology matching process.
  • S-Match: S-Match approach utilizes the semantic information implicitly or explicitly codified in the labels of nodes and arc for computing the semantic correspondences.
  • Alignment-API: Alignment API provides a framework for expressing and sharing ontology alignments. Please note we utilized wordnet based method of Alignment API for matching. Alignment API should be considered as a straw man approach for the purpose of this evaluation.

Comparison Ontology Alignment Evaluation Initiative Oriented Track

Results Ontology Alignment Initiative Oriented Matching Track

System A-API OMViaUO S-Match AROMA RiMOM BLOOMS
Test Prec Rec Prec Rec Prec Rec Prec Rec Prec Rec Prec Rec
1XX 0 0 0.04 0.06 0.01 0.71 NaN 0 1 1 1 1
4XX 0 0 0.01 0.03 0.05 0.30 0.84 0.08 0.67 0.85 0.54 0.51
3XX 0.01 0.03 0.04 0.047 0.01 0.14 0.74 0.11 0.59 0.81 1 0.84
Avg. 0.00 0.01 0.04 0.04 0.03 0.38 0.63 0.07 0.75 0.88 0.84 0.78

Comparison Ontology Alignment Evaluation Initiative Benchmark Track

Results Ontology Alignment Initiative Benchmark Track

System S-Match OMViaUO A-API BLOOMS AROMA RiMOM
Test Prec Rec Prec Rec Prec Rec Prec Rec Prec Rec Prec Rec
1XX 0.11 1 0.26 0.37 0.59 0.96 0.71 1 1 1 1 1
4XX 0.1 0.2 0.21 0.31 0.3 0.54 0.38 0.49 0.88 0.65 0.89 0.78
3XX 0.1 0.2 0.28 0.28 0.45 0.77 0.62 0.84 0.80 0.76 0.80 0.80
Avg. 0.1 0.46 0.25 0.33 0.45 0.76 0.57 0.78 0.88 0.81 0.89 0.66


Comparison Linked Open Data schema Alignment

System A-API OMViaUO RiMOM S-Match AROMA BLOOMS
Test Prec Rec Prec Rec Prec Rec Prec Rec Prec Rec Prec Rec
Music Ontology, BBC Program 0.4 0 1 0 err err 0.04 0.28 0 0 0.63 0.78
Music Ontology, DBpedia 0 0 0 0 err err 0.08 0.30 0.45 0.01 0.39 0.62
FOAF, DBpedia 0 0 0 0 err err 0.11 0.40 0.33 0.04 0.67 0.73
Geonames, DBpedia 0 0 0 0 err err 0.23 1 0 0 0 0
SIOC, FOAF 0 0 0 0 0.3 0.2 0.52 0.11 0.30 0.20 0.55 0.64
Semantic Web Conf. Ontology, AKT Portal Ontology 0.12 0.05 0.16 0.03 err err 0.06 0.4 0.38 0.03 0.42 0.59
Semantic Web Conf. Ontology, DBpedia 0 0 0 0 err err 0.15 0.50 0.27 0.01 0.70 0.40
Avg. 0.07 0.01 0.17 0 NA NA 0.17 0.43 0.25 0.04 0.48 0.54

Alignment API Linked Open Data schema Alignment Performance

Threshold 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
Test Prec Rec Prec Rec Prec Rec Prec Rec Prec Rec Prec Rec Prec Rec Prec Rec Prec Rec Prec Rec
Music Ontology, BBC Program 0.04 0 0.04 0 0.03 0 0.04 0 0.4 0 0.3 0 0.23 0 0.23 0.1 0.13 0.1 0.1 0.1
Music Ontology, DBpedia 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
FOAF, DBpedia 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Geonames, DBpedia 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
SIOC, FOAF 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Semantic Web Conf. Ontology, AKT Portal Ontology 0.06 0.05 0.08 0.04 0.09 0.04 0.08 0.04 0.12 0.05 0.11 0.04 0.1 0.03 0.03 0.05 0.05 0.01 0.02 0.01
Semantic Web Conf. Ontology, DBpedia 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Avg. 0.01 0.0 0.01 0.0 0.01 0.0 0.01 0.00 0.07 0.00 0.05 0.00 0.04 0.004 0.03 0.02 0.02 0.01 0.01 0.01

BLOOMS Team

Resources for Download

Acknowledgement

This work is funded primarily by NSF Award:IIS-0842129, titled III-SGER: Spatio-Temporal-Thematic Queries of Semantic Web Data: a Study of Expressivity and Efficiency. Pascal Hitzler acknowledges support by the Wright State University Research Council.

References

<references/>