Towards Disambiguating Web Tables

Stefan Zwicklbauer; Christoph Einsiedler; Michael Granitzer; Christin Seifert

Towards Disambiguating Web Tables

Stefan Zwicklbauer, Christoph Einsiedler, Michael Granitzer, Christin Seifert

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

20 Citations (Scopus)

10 Downloads (Pure)

Abstract

Web tables comprise a rich source of factual information. However, without semantic annotation of the tables' content the information is not usable for automatic integration and search. We propose a methodology to annotate table headers with semantic type information based on the content of column's cells. In our experiments on 50 tables we achieved an F1 value of 0.55, where the accuracy greatly varies depending on the used ontology. Moreover, we found that for 94% of maximal F1 score only 20 cells (37%) need to be considered on average. Results suggest that for table disambiguation the choice of the ontology needs to be considered and the data input size can be reduced

Original language	English
Title of host publication	ISWC-PD 2013
Subtitle of host publication	ISWC 2013 Posters & Demonstrations Track
Editors	Eva Blomqvist, Tudor Groza
Publisher	CEUR
Pages	205-208
Number of pages	4
Publication status	Published - 2013
Event	12th International Semantic Web Conference, ISWC 2013 - Melbourne, Australia Duration: 21 Oct 2013 → 25 Oct 2013 Conference number: 12

Conference

Conference	12th International Semantic Web Conference, ISWC 2013
Abbreviated title	ISWC
Country	Australia
City	Melbourne
Period	21/10/13 → 25/10/13

Access to Document

Zwicklbauer2013towardsFinal published version, 236 KB

http://ceur-ws.org/Vol-1035/iswc2013_poster_7.pdf

Fingerprint Dive into the research topics of 'Towards Disambiguating Web Tables'. Together they form a unique fingerprint.

Cite this

@inproceedings{a9956fa914e246f4b8d7ab43b9a9888c,

title = "Towards Disambiguating Web Tables",

abstract = "Web tables comprise a rich source of factual information. However, without semantic annotation of the tables' content the information is not usable for automatic integration and search. We propose a methodology to annotate table headers with semantic type information based on the content of column's cells. In our experiments on 50 tables we achieved an F1 value of 0.55, where the accuracy greatly varies depending on the used ontology. Moreover, we found that for 94% of maximal F1 score only 20 cells (37%) need to be considered on average. Results suggest that for table disambiguation the choice of the ontology needs to be considered and the data input size can be reduced",

author = "Stefan Zwicklbauer and Christoph Einsiedler and Michael Granitzer and Christin Seifert",

year = "2013",

language = "English",

pages = "205--208",

editor = "Eva Blomqvist and Tudor Groza",

booktitle = "ISWC-PD 2013",

publisher = "CEUR",

note = "12th International Semantic Web Conference, ISWC 2013, ISWC ; Conference date: 21-10-2013 Through 25-10-2013",

}

TY - GEN

T1 - Towards Disambiguating Web Tables

AU - Zwicklbauer, Stefan

AU - Einsiedler, Christoph

AU - Granitzer, Michael

AU - Seifert, Christin

N1 - Conference code: 12

PY - 2013

Y1 - 2013

N2 - Web tables comprise a rich source of factual information. However, without semantic annotation of the tables' content the information is not usable for automatic integration and search. We propose a methodology to annotate table headers with semantic type information based on the content of column's cells. In our experiments on 50 tables we achieved an F1 value of 0.55, where the accuracy greatly varies depending on the used ontology. Moreover, we found that for 94% of maximal F1 score only 20 cells (37%) need to be considered on average. Results suggest that for table disambiguation the choice of the ontology needs to be considered and the data input size can be reduced

AB - Web tables comprise a rich source of factual information. However, without semantic annotation of the tables' content the information is not usable for automatic integration and search. We propose a methodology to annotate table headers with semantic type information based on the content of column's cells. In our experiments on 50 tables we achieved an F1 value of 0.55, where the accuracy greatly varies depending on the used ontology. Moreover, we found that for 94% of maximal F1 score only 20 cells (37%) need to be considered on average. Results suggest that for table disambiguation the choice of the ontology needs to be considered and the data input size can be reduced

M3 - Conference contribution

SP - 205

EP - 208

BT - ISWC-PD 2013

A2 - Blomqvist, Eva

A2 - Groza, Tudor

PB - CEUR

T2 - 12th International Semantic Web Conference, ISWC 2013

Y2 - 21 October 2013 through 25 October 2013

ER -