Webology, Volume 5, Number 3, September, 2008

Home Table of Contents Titles & Subject Index Authors Index

Folksonomies, the Web and Search Engines

Guest Editor
Louise Spiteri, Ph.D.
School of Information Management
Dalhousie University, Canada


The aim of this special issue of Webology is to explore developments in the design of folksonomies, knowledge organization systems, and search engines to reflect end user preferences for describing items of interest. Particular emphasis is placed on folksonomies, an area of study that has grown exponentially since the term was first coined by Thomas Vander Wal in 2004: "Folksonomy is the result of personal free tagging of information and objects (anything with a URL) for one's own retrieval. The tagging is done in a social environment (usually shared and open to others). Folksonomy is created from the act of tagging by the person consuming the information" (Vander Wal, 2007). Since 2004, social software applications and their use of tagging have continued to increase in popularity; in its site dedicated to such applications, Wikipedia (2008) lists no less that 11 extant media sharing sites and 26 social bookmarking sites. This list does not take into account the approximate 20 media cataloguing sites, not to mention the innumerable blogging sites that employ tagging.

The state of folksonomy research has similarly experienced a remarkable growth since 2004; when I first started my research in this field in early 2005, it was difficult to find much literature beyond the blogosphere. As of this writing, my "folksonomy bibliography" features more than 300 references, many of which are from scholarly conferences and peer-reviewed academic journals. Folksonomy research is conducted in a wide variety of disciplines and fields of study, including:

Articles in This Issue

The papers in this special issue reflect the diversity of approaches taken to create Web resources that reflect better the needs of end users. Particular emphasis is placed on the need to manage the increasing volumes of tags and information available on the Web, particularly as more people are becoming engaged with numerous social applications. As is discussed in some of the papers in this special edition, there is certainly scope to consider ways in which to combine the more traditional controlled vocabularies with the free-flowing nature of tagging.

Isabella Peters and Katrin Weller apply gardening as a metaphor to describe a method by which end users can manage folksonomies to assist them in improving the efficiency and relevance of their search efforts. Peters and Weller argue that end users often have a collection of tagged items scattered across the different social applications to which they belong. Using the metaphor of a well-maintained garden that is cultivated, pruned, and weeded, Peters and Weller suggest that end users need a tool to help them collect and organize all these scattered tags in a centralized location. To this end, Peters and Weller have created TagCare, which allows its members to import their tags from Flickr, Bibsonomy, or Del.icio.us. , and to create their own personomy (akin to a thesaurus), in which they can determine the hierarchical, equivalence, and semantic relationships amongst tags. TagCare may serve as an important tool by which to create a bridge between uncontrolled folksonomies and controlled vocabularies, particularly if, as they hope, their system can be extended to creating community gardens of controlled folksonomies.

Robert Bruce investigates the relationship between tags and controlled vocabularies by examining the overlap of tags assigned to articles in CiteULike and in the Education Resources Information Center (ERIC) database. Bruce compared tags assigned in CiteULike, a social citation service that allows members to manage and tag their journal citations, to the ERIC descriptors assigned to a set of 2,786 journal articles.
Bruce's findings suggest that little overlap (7.6%) exists between the descriptors assigned by subject specialists in ERIC and the tags used by CiteULike users. It would be tempting to conclude that the ERIC thesaurus is not very reflective of the language of CiteULike members, but Bruce admits to important limitations in his study, namely that he examined only exact matches and did not take into account variant spellings, spelling mistakes, and grammatical variants (e.g., nouns v. verbs). Bruce suggests that further research that includes a semantic analysis of the two sets of terms may lead to a greater understanding of the nature of the two vocabularies and of how CiteULike tags could be incorporated in the ERIC thesaurus.

Maja van der Velden examines how development information (i.e., concerning developing nations) on the Internet is organized and whether the classification schemes used reflect the cultural diversity and needs of the people who are most likely to need and use this information. Van der Velden maps 7 categories of need expressed by the Declaration of Concerned Knowledge Workers to 11 Web resources to look for similarities and differences in the way these resources classify development knowledge. The results of this mapping suggest that the 11 Web resources differ widely in how they classify development knowledge and that the categories they use are controlled by centralized editorial polices that may exclude the knowledge communities that use the information. Van der Velden suggests that an analysis of end-user tagging could be used to build inclusive classification systems for Web-based development knowledge.

Chihli Hung addresses the problems that online consumers often face in finding credible product reviews to assist them in their purchases. Hung's emphasis is upon consumer-to-consumer product reviews, or what he refers to as word-of-mouth reviews, which may often be more useful and critical than those found in professional consumer reports. Hung proposes a conceptual model that mines personalized word-of-mouth information. Hung's model analyzes the language used by consumers to describe a product and ranks the products based on consumer preferences. The model's output is a list of recommendations (either for or against) for a given product, which could serve to greatly aid consumers in their purchasing decisions.


Bibliographic information of this paper for citing:

Spiteri, Louise (2008).   "Editorial: Folksonomies, the Web and Search Engines."   Webology, 5(3), editorial 17. Available at: http://www.webology.org/2008/v5n3/editorial17.html

Copyright © 2008, Louise Spiteri.