Details
Originalsprache | Englisch |
---|---|
Titel des Sammelwerks | Advanced Data Mining and Applications |
Untertitel | 5th International Conference, ADMA 2009, Proceedings |
Seiten | 31-42 |
Seitenumfang | 12 |
ISBN (elektronisch) | 978-3-642-03348-3 |
Publikationsstatus | Veröffentlicht - 2009 |
Veranstaltung | 5th International Conference on Advanced Data Mining and Applications, ADMA 2009 - Beijing, China Dauer: 17 Aug. 2009 → 19 Aug. 2009 |
Publikationsreihe
Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
---|---|
Band | 5678 LNAI |
ISSN (Print) | 0302-9743 |
ISSN (elektronisch) | 1611-3349 |
Abstract
Web 2.0 applications such as delicious, flickr or lastfm have recently become extremely popular and as a result, a large amount of semantically rich metadata produced by users becomes available and exploitable. Tag information can be used for many purposes (e.g. user profiling, recommendations, clustering etc), though the benefit of tags for search is by far the most discussed usage. Tag types differ largely across systems and previous studies showed that, while some tag type categories might be useful for some particular users when searching, they may not bring any benefit to others. The present paper proposes an approach which utilizes rule-based as well as model-based methods, in order to automatically identify exactly these different types of tags. We compare the automatic tag classification produced by our algorithms against a ground truth data set, consisting of manual tag type assignments produced by human raters. Experimental results show that our methods can identify tag types with high accuracy, thus enabling further improvement of systems making use of social tags.
ASJC Scopus Sachgebiete
- Mathematik (insg.)
- Theoretische Informatik
- Informatik (insg.)
Zitieren
- Standard
- Harvard
- Apa
- Vancouver
- BibTex
- RIS
Advanced Data Mining and Applications: 5th International Conference, ADMA 2009, Proceedings. 2009. S. 31-42 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Band 5678 LNAI).
Publikation: Beitrag in Buch/Bericht/Sammelwerk/Konferenzband › Aufsatz in Konferenzband › Forschung › Peer-Review
}
TY - GEN
T1 - Automatically identifying tag types
AU - Bischoff, Kerstin
AU - Firan, Claudiu S.
AU - Kadar, Cristina
AU - Nejdl, Wolfgang
AU - Paiu, Raluca
PY - 2009
Y1 - 2009
N2 - Web 2.0 applications such as delicious, flickr or lastfm have recently become extremely popular and as a result, a large amount of semantically rich metadata produced by users becomes available and exploitable. Tag information can be used for many purposes (e.g. user profiling, recommendations, clustering etc), though the benefit of tags for search is by far the most discussed usage. Tag types differ largely across systems and previous studies showed that, while some tag type categories might be useful for some particular users when searching, they may not bring any benefit to others. The present paper proposes an approach which utilizes rule-based as well as model-based methods, in order to automatically identify exactly these different types of tags. We compare the automatic tag classification produced by our algorithms against a ground truth data set, consisting of manual tag type assignments produced by human raters. Experimental results show that our methods can identify tag types with high accuracy, thus enabling further improvement of systems making use of social tags.
AB - Web 2.0 applications such as delicious, flickr or lastfm have recently become extremely popular and as a result, a large amount of semantically rich metadata produced by users becomes available and exploitable. Tag information can be used for many purposes (e.g. user profiling, recommendations, clustering etc), though the benefit of tags for search is by far the most discussed usage. Tag types differ largely across systems and previous studies showed that, while some tag type categories might be useful for some particular users when searching, they may not bring any benefit to others. The present paper proposes an approach which utilizes rule-based as well as model-based methods, in order to automatically identify exactly these different types of tags. We compare the automatic tag classification produced by our algorithms against a ground truth data set, consisting of manual tag type assignments produced by human raters. Experimental results show that our methods can identify tag types with high accuracy, thus enabling further improvement of systems making use of social tags.
KW - Classification
KW - Collaborative tagging
KW - Social media
KW - Tag types
UR - http://www.scopus.com/inward/record.url?scp=70350340351&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-03348-3_7
DO - 10.1007/978-3-642-03348-3_7
M3 - Conference contribution
AN - SCOPUS:70350340351
SN - 978-3-642-03347-6
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 31
EP - 42
BT - Advanced Data Mining and Applications
T2 - 5th International Conference on Advanced Data Mining and Applications, ADMA 2009
Y2 - 17 August 2009 through 19 August 2009
ER -