Graduate School and Research Center in Digital Sciences

Improving schema matching with linked data

Assaf, Ahmad; Louw, Eldad; Senart, Aline; Follenfant, Corentin; Troncy, Raphaël; Trastour, David

WOD 2012, 1st International Workshop on Open Data, May 25th, 2012, Nantes, France / Also published on ArXiv

With today's public data sets containing billions of data items, more and more companies are looking to integrate external data with their traditional enterprise data to improve business intelligence analysis. These distributed data sources however exhibit heterogeneous data formats and terminologies and may contain noisy data. In this paper, we present a novel framework that enables business users to semi-automatically perform data integration on potentially noisy tabular data. This framework o ers an extension to Google Re ne with novel schema matching algorithms leveraging Freebase rich types. First experiments show that using Linked Data to map cell values with instances and column headers with types improves signi cantly the quality of the matching results and therefore should lead to more informed decisions.

Document Arxiv Bibtex

Title:Improving schema matching with linked data
Type:Conference
Language:English
City:Nantes
Country:FRANCE
Date:
Department:Data Science
Eurecom ref:4447
Copyright: © ACM, 2012. This is the author's version of the work. It is posted here by permission of ACM for your personal use. Not for redistribution. The definitive version was published in WOD 2012, 1st International Workshop on Open Data, May 25th, 2012, Nantes, France / Also published on ArXiv
Bibtex: @inproceedings{EURECOM+4447, year = {2012}, title = {{I}mproving schema matching with linked data}, author = {{A}ssaf, {A}hmad and {L}ouw, {E}ldad and {S}enart, {A}line and {F}ollenfant, {C}orentin and {T}roncy, {R}apha{\"e}l and {T}rastour, {D}avid}, booktitle = {{WOD} 2012, 1st {I}nternational {W}orkshop on {O}pen {D}ata, {M}ay 25th, 2012, {N}antes, {F}rance / {A}lso published on {A}r{X}iv }, address = {{N}antes, {FRANCE}}, month = {05}, url = {http://www.eurecom.fr/publication/4447} }
See also: