Ecole d'ingénieur et centre de recherche en Sciences du numérique

ADEL@OKE 2017: A generic method for indexing knowlege bases for entity linking

Plu, Julien; Troncy, Raphaël; Rizzo, Giuseppe

ESWC 2017, 14th European Semantic Web Conference, Open Extraction Challenge, 28th May-1st June 2017, Portoroz, Slovenia

Open Knowledge Extraction Challenge Award

In this paper, we report on the participation of ADEL, an adaptive entity recognition and linking framework, to the OKE 2017 challenge. In particular, we propose an hybrid approach that combines various extraction methods to improve the recognition level and an efficient knowledge base indexing process to increase the efficiency of the linking step. We detail how we deal with finegrained entity types, either generic (e.g. Activity, Competition, Animal for the task 2) or domain specific (e.g. MusicArtist, SignalGroup, MusicalWork for the task 3). We also show how ADEL can flexibly disambiguate entities from different knowledge bases (DBpedia and MusicBrainz). We obtain promising results on the OKE 2017 challenge training dataset for the first three tasks.

Document Bibtex

Titre:ADEL@OKE 2017: A generic method for indexing knowlege bases for entity linking
Mots Clés:Entity Recognition, Entity Linking, Feature Extraction, Indexing, OKE Challenge, ADEL
Type:Conférence
Langue:English
Ville:Portoroz
Pays:SLOVÉNIE
Date:
Département:Data Science
Eurecom ref:5231
Copyright: © Springer. Personal use of this material is permitted. The definitive version of this paper was published in ESWC 2017, 14th European Semantic Web Conference, Open Extraction Challenge, 28th May-1st June 2017, Portoroz, Slovenia and is available at :
Bibtex: @inproceedings{EURECOM+5231, year = {2017}, title = {{ADEL}@{OKE} 2017: {A} generic method for indexing knowlege bases for entity linking}, author = {{P}lu, {J}ulien and {T}roncy, {R}apha{\"e}l and {R}izzo, {G}iuseppe }, booktitle = {{ESWC} 2017, 14th {E}uropean {S}emantic {W}eb {C}onference, {O}pen {E}xtraction {C}hallenge, 28th {M}ay-1st {J}une 2017, {P}ortoroz, {S}lovenia}, address = {{P}ortoroz, {SLOV}{\'{E}}{NIE}}, month = {05}, url = {http://www.eurecom.fr/publication/5231} }
Voir aussi: