Towards Ontological Structures Extraction from Folksonomies: An Efficient Fuzzy Clustering Approach

Towards Ontological Structures Extraction from Folksonomies: An Efficient Fuzzy Clustering Approach

Marouf Zahia, Benslimane Sidi Mohamed
Copyright: © 2014 |Pages: 11
DOI: 10.4018/ijiit.2014100103
(Individual Articles)
No Current Special Offers


Folksonomies are one of the technologies of Web 2.0 that permit users to annotate resources on the Web. In this paper, the authors propose an integrated approach to extract ontological structures from unstructured and semi-structured resources. Our proposal overcome limitations of existing approaches. It gives a formal, simple, and efficient solution to the tag clustering and disambiguation problem. Moreover, their approach doesn't need any ontology as an upper guide during the generation process. The generated ontology can be used to enhance various tasks such as ontology evolution and enrichment.
Article Preview


Folksonomies (Vander, 2007) have recently emerged as a powerful way to label and organize large collections of data. These systems allow users to use any keywords or tags relevant to the content to annotate their favorite resources on the web. In this paper, we propose an approach to creating a hierarchy that captures the hierarchical semantic structure of folksonomies. Our approach combines the advantages of folksonomies and ontologies, and overcomes the ontology learning difficulties and costs. The main contributions of the present work are:

  • A new representation of the folksonomy based on a tag-tag binary matrix.

  • A new similarity measure that calculates relatedness between tags.

  • A new generality measure that determines the abstractness level of tags.

  • We address the issue of ambiguity in folksonomy based services to return more results that would be missed by traditional search mechanisms.

  • A new fuzzy clustering algorithm that doesn’t need determining previously the number of cluster, but deduces it when performing the clustering process.

  • Finally, we conduct comprehensive experiments to evaluate our approach by using real dataset from social bookmarking systems.

The rest of the paper is structured as follows. In section 2, we provide an overview of related work. In section 3, we outline the proposed approach Section 4 introduces an experimental methodology to evaluate the approach. Finally, section 5 concludes the paper and points directions for future work.


Existing approaches to infer tag relationships can broadly be assigned to one of the following classes:

  • Clustering approaches: These approaches identify the semantics of tags by clustering tags according to some relations among them (Mika, 2007; Hamasaki, 2007; Begelman, 2006; Kennedy, 2007; Heymann, 2006; Benz, 2010; Marouf, 2013). These clustering approaches measure tags similarity based only on one dimension (user or resource), and most of them do not deal with ambiguity problem or do not give a formal solution to it. Furthermore, the majority of these approaches do not make the hierarchical relations explicit between tags.

  • Association rule mining approaches: This class of approaches applies association rule mining techniques in the ontology learning process (Schmitz, 2006; Jäschke, 2008; Trabelsi, 2012). These approaches output a hierarchical representation of tags described by a tri-ordered set of triples called triadic concepts in Formal Concept Analysis FCA (Lehmann, 1995), where each triple consists of a set of users, a set of tags, and a set of resources. However, the relationships between tags in different hierarchical levels are not defined semantically, and there is no strategy to deal with ambiguous tags.

  • Semantic based approaches: These approaches aim at associating semantic entities to tags as a way to formally define their meaning (Angeletou, 2008; Cantador, 2008; Garcia, 2009; Djuana, 2012). The lack of ontologies that well match the tags in folksonomies is one of the major obstacles applying these approaches.

  • Hybrid approaches: these approaches integrate multiple techniques from those cited previously (Giannakidou, 2008; Specia, 2007; Lin, 2009; Schmitz, 2006). These approaches are however limited as they do not present any disambiguation activity.

Complete Article List

Search this Journal:
Volume 20: 1 Issue (2024)
Volume 19: 1 Issue (2023)
Volume 18: 4 Issues (2022): 3 Released, 1 Forthcoming
Volume 17: 4 Issues (2021)
Volume 16: 4 Issues (2020)
Volume 15: 4 Issues (2019)
Volume 14: 4 Issues (2018)
Volume 13: 4 Issues (2017)
Volume 12: 4 Issues (2016)
Volume 11: 4 Issues (2015)
Volume 10: 4 Issues (2014)
Volume 9: 4 Issues (2013)
Volume 8: 4 Issues (2012)
Volume 7: 4 Issues (2011)
Volume 6: 4 Issues (2010)
Volume 5: 4 Issues (2009)
Volume 4: 4 Issues (2008)
Volume 3: 4 Issues (2007)
Volume 2: 4 Issues (2006)
Volume 1: 4 Issues (2005)
View Complete Journal Contents Listing