First Insights into the Library Track of the OAEI Dominique Ritze Mannheim University Library
Motivation Ontology Mapping Search Publication x 0 results subject (thesaurus 2): ontology alignment Thesaurus 1 Ontology Mapping Thesaurus 2 = Ontology Mapping Search Ontology Alignment Publication x subject (thesaurus 1): ontology alignment
Overview • Ontology Matching • OAEI • Thesaurus vs. Ontology • OAEI Library Track 2012 • Lessons learned and Future Work
Ontology Matching Person People Author Committee. Member PCMember Author < Author, =, 0. 97 > < Paper, =, 0. 94 > < reviews, =, 0. 91 > < writes, =, 0. 7 > < Person, People, =, 0. 8 > < Document, Doc, =, 0. 7 > < Reviewer, Review, =, 0. 6 > writes Reviewer reviews … reviews Document Paper Review Doc Paper writes
Ontology Matching Evaluation O 1 Tool Test A O 2 m R Result
Ontology Alignment Evaluation Initiative (OAEI) Annual campaign started 2005 Different tracks/datasets • Benchmark, Anatomy, Conference, Multifarm, Large Bio. Med, Library, Instance Matching • 21 submitted systems (2012) • Goal: Improving the performances of the ontology matching field • Through comparison of algorithms • New challenges for the systems • •
Thesaurus = Ontology? SKOS OWL Germany skos: concept skos: pref. Label skos: alternative. Label Commodities Tropical Fruit owl: class rdfs: label Ananas skos: scope. Note rdfs: comment skos: notation. Metal Product -> Metal A skos: narrower B A rdfs: sub. Class. Of B A skos: broader B B rdfs: sub. Class. Of A skos: related rdfs: see. Also
OAEI Library Track Are current state-of-the-art ontology matching tools able to match thesauri? Dominique Ritze, Kai Eckert, Benjamin Zapilko, Joachim Neubert
Data Set Thesaurus for economics (STW) • 6. 000 concepts with 19. 000 additional keywords (EN, DE) • Thesaurus for the Sociel Sciences (The. Soz) • 8. 000 concepts with 4. 000 additional keywords (EN, DE, FR) • • • Reference alignment manually created in 2006 Both actively used in libraries for keyword indexing
Execution • • • 7 GB Debian machine Timeframe 1 week 13 of the 21 submitted systems were able to generate an alignment No system had a heap space problem Evaluation: Precision, Recall, F-Measure, Runtime
Results System GOMMA Serv. OMap. Lt Log. Map Serv. OMap YAM++ Log. Map. Lt G 02 A Hertuda We. Se. E Hot. Match CODI Map. SSS AROMA Optima Precision Recall F-Measure 0. 537 0. 906 0. 674 0. 654 0. 687 0. 670 0. 688 0. 644 0. 665 0. 717 0. 619 0. 665 0. 595 0. 750 0. 664 0. 577 0. 776 0. 662 0. 675 0. 645 0. 660 0. 465 0. 925 0. 619 0. 612 to evaluate the results? F 0. 607 0. 609 How 0. 645 Measure of 0. 575 0. 67 good? 0. 608 0. 434 0. 481 0. 456 0. 520 0. 184 0. 272 0. 107 0. 652 0. 184 0. 321 0. 072 0. 117 Time (s) Size 804 45 95 44 496 21 32773 14363 144070 14494 39869 2171 1096 37457 4712 2938 2620 2413 3522 3756 2671 5559 2774 2494 3100 989 17001 624 1: 1 yes yes yes
Results System Precision Recall F-Measure Time (s) Size Matcher. Pref Matcher. DE Matcher. All GOMMA Serv. OMap. Lt Log. Map 0. 820 0. 891 0. 544 0. 537 0. 654 0. 688 0. 642 0. 601 0. 896 0. 906 0. 687 0. 644 0. 720 0. 717 0. 674 0. 670 0. 665 75 42 735 804 45 95 2190 1885 4605 4712 2938 2620 0. 569 0. 456 0. 272 0. 184 0. 117 36 39869 2171 1096 37457 1518 3100 989 17001 624 1: 1 … Matcher. EN CODI Map. SSS AROMA Optima 0. 808 0. 434 0. 520 0. 107 0. 321 0. 439 0. 481 0. 184 0. 652 0. 072 yes
Manual Evaluation • • Between 38 and 269 new correct correspondences found per matcher Up to half of the correspondences correct Many new correspondences are quite simple Some more “complex” and interesting ones • Automated production = CAM • Several incorrect ones if the labels are quite similar • Difficult to distinguish the names of countries, their inhabitants and the languages
Lessons Learned • • • Transformation SKOS to OWL causes some problems, especially regarding the labels Ontology matching systems are nevertheless able to match thesauri and even discover unknown correct correspondences Interest of the community in this topic
Future Work • Update reference alignment adapted results • • • SKOS import for matching systems Use instance data to match thesauri? Other thesauri?
Thank you for your attention! dominique. ritze@bib. uni-mannheim. de
Скачать презентацию First Insights into the Library Track of the
f197f795688462ccb725063f8a2943ad.ppt