Ranking social bookmarks using topic models

M. Harvey, I. Ruthven, M. Carman

Research output: Contribution to conferencePaper

4 Citations (Scopus)

Abstract

Ranking of resources in social tagging systems is a difficult problem due to the inherent sparsity of the data and the vo- cabulary problems introduced by having a completely unre- stricted lexicon. In this paper we propose to use hidden topic models as a principled way of reducing the dimensionality of this data to provide more accurate resource rankings with higher recall. We first describe Latent Dirichlet Allocation (LDA) and then show how it can be used to rank resources in a social bookmarking system. We test the LDA tagging model and compare it with 3 non-topic model baselines on a large data sample obtained from the Delicious social book- marking site. Our evaluations show that our LDA-based method significantly outperforms all of the baselines.

Conference

Conference19th ACM international conference on Information and knowledge management
CountryCanada
CityToronto
Period26/10/1030/10/10

Keywords

  • social bookmarking
  • social tagging
  • ranking
  • latent dirichlet allocation
  • topic models
  • social bookmarks

Cite this

Harvey, M., Ruthven, I., & Carman, M. (2010). Ranking social bookmarks using topic models. 1401-1404 . Paper presented at 19th ACM international conference on Information and knowledge management , Toronto, Canada. https://doi.org/10.1145/1871437.1871632
Harvey, M. ; Ruthven, I. ; Carman, M. / Ranking social bookmarks using topic models. Paper presented at 19th ACM international conference on Information and knowledge management , Toronto, Canada.4 p.
@conference{c95d463f77d34ba59470b3195b4f0fa1,
title = "Ranking social bookmarks using topic models",
abstract = "Ranking of resources in social tagging systems is a difficult problem due to the inherent sparsity of the data and the vo- cabulary problems introduced by having a completely unre- stricted lexicon. In this paper we propose to use hidden topic models as a principled way of reducing the dimensionality of this data to provide more accurate resource rankings with higher recall. We first describe Latent Dirichlet Allocation (LDA) and then show how it can be used to rank resources in a social bookmarking system. We test the LDA tagging model and compare it with 3 non-topic model baselines on a large data sample obtained from the Delicious social book- marking site. Our evaluations show that our LDA-based method significantly outperforms all of the baselines.",
keywords = "social bookmarking, social tagging, ranking, latent dirichlet allocation, topic models, social bookmarks",
author = "M. Harvey and I. Ruthven and M. Carman",
year = "2010",
month = "10",
day = "1",
doi = "10.1145/1871437.1871632",
language = "English",
pages = "1401--1404",
note = "19th ACM international conference on Information and knowledge management ; Conference date: 26-10-2010 Through 30-10-2010",

}

Harvey, M, Ruthven, I & Carman, M 2010, 'Ranking social bookmarks using topic models' Paper presented at 19th ACM international conference on Information and knowledge management , Toronto, Canada, 26/10/10 - 30/10/10, pp. 1401-1404 . https://doi.org/10.1145/1871437.1871632

Ranking social bookmarks using topic models. / Harvey, M.; Ruthven, I.; Carman, M.

2010. 1401-1404 Paper presented at 19th ACM international conference on Information and knowledge management , Toronto, Canada.

Research output: Contribution to conferencePaper

TY - CONF

T1 - Ranking social bookmarks using topic models

AU - Harvey, M.

AU - Ruthven, I.

AU - Carman, M.

PY - 2010/10/1

Y1 - 2010/10/1

N2 - Ranking of resources in social tagging systems is a difficult problem due to the inherent sparsity of the data and the vo- cabulary problems introduced by having a completely unre- stricted lexicon. In this paper we propose to use hidden topic models as a principled way of reducing the dimensionality of this data to provide more accurate resource rankings with higher recall. We first describe Latent Dirichlet Allocation (LDA) and then show how it can be used to rank resources in a social bookmarking system. We test the LDA tagging model and compare it with 3 non-topic model baselines on a large data sample obtained from the Delicious social book- marking site. Our evaluations show that our LDA-based method significantly outperforms all of the baselines.

AB - Ranking of resources in social tagging systems is a difficult problem due to the inherent sparsity of the data and the vo- cabulary problems introduced by having a completely unre- stricted lexicon. In this paper we propose to use hidden topic models as a principled way of reducing the dimensionality of this data to provide more accurate resource rankings with higher recall. We first describe Latent Dirichlet Allocation (LDA) and then show how it can be used to rank resources in a social bookmarking system. We test the LDA tagging model and compare it with 3 non-topic model baselines on a large data sample obtained from the Delicious social book- marking site. Our evaluations show that our LDA-based method significantly outperforms all of the baselines.

KW - social bookmarking

KW - social tagging

KW - ranking

KW - latent dirichlet allocation

KW - topic models

KW - social bookmarks

UR - http://bradipo.net/mark/papers/cikm2010_p1401-harvey.pdf

UR - http://www.cis.strath.ac.uk/cis/research/publications/papers/strath_cis_publication_2478.pdf

UR - http://www.yorku.ca/cikm10/

U2 - 10.1145/1871437.1871632

DO - 10.1145/1871437.1871632

M3 - Paper

SP - 1401

EP - 1404

ER -

Harvey M, Ruthven I, Carman M. Ranking social bookmarks using topic models. 2010. Paper presented at 19th ACM international conference on Information and knowledge management , Toronto, Canada. https://doi.org/10.1145/1871437.1871632