Exploit the tripartite network of social tagging for web clustering

Lu, Caimei; Chen, Xin; Park, E. K.

doi:10.1145/1645953.1646167

Cited by 30 publications

(29 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…As the results in the table indicate, inclusion of tag information in any form seems to improve the performance as compared to the case when only words from page-text are used. This is evidenced by the better results of words + tags as compared to words only and tags only (which has also been shown in some other recent works [34,29]). …”

Section: Full Datasupporting

confidence: 74%

“…Other recent works on exploiting social annotations, in particular, to improve various web mining tasks include annotation based approaches to web search [4], webpage classification [45], and information retrieval in general [44]. Similar in spirit to our work, using tag information for webpage clustering has earlier been proposed in [34,29] using a concatenation of word and tag feature vectors. In [34], the authors also proposed a probabilistic generative model based on an extension of the Latent Dirichlet Allocation [9].…”

Section: Related Workmentioning

confidence: 75%

“…To assess the relevance and coherency of the discovered clusters, one can use hierarchical web directories such as the Open Directory Project [1] (ODP) as the gold standard. Web directories such as ODP are widely acceptable gold standards because they usually provide an agreed-upon clustering of webpages by human users, and have been used for evaluations in various recent works [34,29].…”

Section: Webpage Clustering Using Tagsmentioning

confidence: 99%

“…Some of the common choices are [34,29] It turns out [34,29] that the concatenation of word and tag feature vectors (4) outperforms approaches that use feature vectors derived from the word (1) vocabulary, the tag vocabulary (2), or vocabulary derived from a union of words and tags (3).…”

Section: Webpage Clustering Using Tagsmentioning

confidence: 99%

“…To our knowledge, all the existing approaches exploiting tag information for webpage clustering (e.g., [34,29] and the CCA based approach we proposed in this paper) assume that all the webpages are tagged, which is a somewhat restrictive assumption. In a more realistic setting, one can only expect that the tags will be available for only a small number of webpages.…”

Section: Future Workmentioning

confidence: 99%

See 4 more Smart Citations

Exploiting tag and word correlations for improved webpage clustering

Trivedi

Rai

DuVall

et al. 2010

Proceedings of the 2nd International Workshop on Search and Mining User-Generated Contents

View full text Add to dashboard Cite

Automatic clustering of webpages helps a number of information retrieval tasks, such as improving user interfaces, collection clustering, introducing diversity in search results, etc. Typically, webpage clustering algorithms only use features extracted from the page-text. However, the advent of social-bookmarking websites, such as StumbleUpon 1 and Delicious 2 , has led to a huge amount of user-generated content such as the tag information that is associated with the webpages. In this paper, we present a subspace based feature extraction approach which leverages tag information to complement the page-contents of a webpage to extract highly discriminative features, with the goal of improved clustering performance. In our approach, we consider page-text and tags as two separate views of the data, and learn a shared subspace that maximizes the correlation between the two views. Any clustering algorithm can then be applied in this subspace. We compare our subspace based approach with a number of baselines that use tag information in various other ways, and show that the subspace based approach leads to improved performance on the webpage clustering task. Although our results here are on the webpage clustering task, the same approach can be used for webpage classification as well. In the end, we also suggest possible future work for leveraging tag information in webpage clustering, especially when tag information is present for not all, but only for a small number of webpages. * Also holds an adjunct position with the

show abstract

Section: Full Datasupporting

confidence: 74%