We describe two variants of our approach to tackle the task 1 & 2 of the ECML PKDD Discovery Challenge 2009 where each contenter had to identify up to 5 tags for each resource of a given set of either bibtex-like references to publications or bookmarks. The quality of the results was measured against the tags that users of the data source (www.bibsonomy.org) had originally assigned to the resources (F1 measure). In our approach, we either generate tags (from the content of the given resource data or after crawling additional resources) or we request tags from tagging services. We call each of this tag sources a tag recommender. We then combine the results of the tag recommenders based on weighting factors. The weighting factors are determined experimentally by comparing generated and expected tags based on the available training data. This general idea is also used for the graph-based approach required to solve task 2. Here again, the final tag recommendations are computed from the individual results of the different tag-recommending algorithms. In the preliminary result list, we ranked second for task 1 (Group 2) and nineth for task 2 (Group 1).
%0 Conference Paper
%1 marinho:ecml2009
%A Mrosek, Johannes
%A Bussmann, Stefan
%A Albers, Hendrik
%A Posdziech, Kai
%A Hengefeld, Benedikt
%A Opperman, Nils
%A Robert, Stefan
%A Spira, Gerrit
%B ECML PKDD Discovery Challenge 2009 (DC09)
%C Bled, Slovenia
%D 2009
%E Eisterlehner, Folke
%E Hotho, Andreas
%E Jäschke, Robert
%I CEUR Workshop Proceedings
%K 2009 BibSonomy ECML09 _todo content graph recommendation tagging
%P 189--199
%T Content- and Graph-based Tag Recommendation: Two Variations
%U http://sunsite.informatik.rwth-aachen.de/Publications/CEUR-WS/Vol-497/
%V 497
%X We describe two variants of our approach to tackle the task 1 & 2 of the ECML PKDD Discovery Challenge 2009 where each contenter had to identify up to 5 tags for each resource of a given set of either bibtex-like references to publications or bookmarks. The quality of the results was measured against the tags that users of the data source (www.bibsonomy.org) had originally assigned to the resources (F1 measure). In our approach, we either generate tags (from the content of the given resource data or after crawling additional resources) or we request tags from tagging services. We call each of this tag sources a tag recommender. We then combine the results of the tag recommenders based on weighting factors. The weighting factors are determined experimentally by comparing generated and expected tags based on the available training data. This general idea is also used for the graph-based approach required to solve task 2. Here again, the final tag recommendations are computed from the individual results of the different tag-recommending algorithms. In the preliminary result list, we ranked second for task 1 (Group 2) and nineth for task 2 (Group 1).
@inproceedings{marinho:ecml2009,
abstract = {We describe two variants of our approach to tackle the task 1 & 2 of the ECML PKDD Discovery Challenge 2009 where each contenter had to identify up to 5 tags for each resource of a given set of either bibtex-like references to publications or bookmarks. The quality of the results was measured against the tags that users of the data source (www.bibsonomy.org) had originally assigned to the resources (F1 measure). In our approach, we either generate tags (from the content of the given resource data or after crawling additional resources) or we request tags from tagging services. We call each of this tag sources a tag recommender. We then combine the results of the tag recommenders based on weighting factors. The weighting factors are determined experimentally by comparing generated and expected tags based on the available training data. This general idea is also used for the graph-based approach required to solve task 2. Here again, the final tag recommendations are computed from the individual results of the different tag-recommending algorithms. In the preliminary result list, we ranked second for task 1 (Group 2) and nineth for task 2 (Group 1).},
added-at = {2010-01-29T16:41:25.000+0100},
address = {Bled, Slovenia},
author = {Mrosek, Johannes and Bussmann, Stefan and Albers, Hendrik and Posdziech, Kai and Hengefeld, Benedikt and Opperman, Nils and Robert, Stefan and Spira, Gerrit},
biburl = {https://www.bibsonomy.org/bibtex/220415a7967ab1330d2226b176e11a8d9/trude},
booktitle = {ECML PKDD Discovery Challenge 2009 (DC09)},
editor = {Eisterlehner, Folke and Hotho, Andreas and Jäschke, Robert},
interhash = {c7e9c420a33ec61179eae811d4217b4c},
intrahash = {20415a7967ab1330d2226b176e11a8d9},
issn = {1613-0073},
keywords = {2009 BibSonomy ECML09 _todo content graph recommendation tagging},
month = {September},
pages = {189--199},
publisher = {CEUR Workshop Proceedings},
timestamp = {2010-01-29T16:41:25.000+0100},
title = {Content- and Graph-based Tag Recommendation: Two Variations},
url = {http://sunsite.informatik.rwth-aachen.de/Publications/CEUR-WS/Vol-497/},
volume = 497,
year = 2009
}