Citation Context

... similar topics (e.g. San Francisco), instead of trying to identify keywords which are most important to a document. It is found that tags are more likely to be broad terms rather than specific terms =-=[25]-=-, suggesting that tags are more likely to be used to categorise a document. Hence, by considering the frequencies of tags in term vectors, we should be able to group documents into different categorie...

Citation Context

...m incoming hyperlinks (by web authors) and document popularity. We included document popularity because we found that user behavior for social bookmarking varies considerably with document popularity =-=[19, 21]-=-. For the social annotation dimension, we used del.icio.us, the main social bookmarking site in the WWW today. The Yahoo! company has a large user base with more than 2 million registered users in 200...

Citation Context

... a different task: social classification. 3.2 Classification There is little work dealing with the analysis of the usefulness of social tags for classification tasks. An early work by Noll and Meinel =-=[18]-=- presents a study of the characteristics of social annotations provided by end users, in order to determine their usefulness for web page classification. The authors matched user-supplied tags of a pa...

Citation Context

...ide a very quick impression of trends or “hot topics”. Research on tag clouds is relatively scarce. Publications in this context often merely use the tags as a resource for data-mining tasks (e.g. in =-=[6]-=-). The visual effects used in tag clouds were analysed by Bateman, Gutwin and Necenta [7]. They found that font-size, font-weight and intensity had the strongest visual influence on a user’s perceptio...

Citation Context

...ther hand, researchers investigated to what extent tags (and the resources they annotate) can be semantically grounded and classified into predefined semantic categories. For example, Noll and Meinel =-=[8]-=- presented a study of the characteristics of tags and determined their usefulness for web page classification [9]. Similar to our work, Overell et al. [10] presented an approach which allows classifyi...

Citation Context

...nt of conference. K. M. Schneider [15] has extracted information of conferences from CFPs. In recent years, exploring social tags for classification has attracted considerable attentions. Noll et al. =-=[16]-=- systematically have analyzed exploring social tags for web document classification. Sakurai et al. [17] have proposed a method that classifies bloggers using social tags. Yin et al. [18] have explore...

Citation Context

...mmers may post to the site, their posts will not be displayed to data consumers unless the data consumers are searching for information about those spammers. Other researchers [11], [12], [13], [14], =-=[15]-=- explore annotations as a medium for optimizing Web searches and for classification tasks. The primary difference between these approaches and PASM is the audience. PASM is meant for users to annotate...

Citation Context

... of the distillation of broad search topics through the discovery of “authoritative” information sources on these topics.They proposed and tested an algorithmic formulation of the notion of authority The Scientific World Journal 3 that is based on the relationship between a set of relevant authoritative pages and the set of “hub pages” that join them together in the link structure.Their formulation is associated with the eigenvectors of certain matrices associated with the link graph. These associations in turn motivate additional heuristics for link-based analysis. In Noll and Meinel’s study [11], social annotation via the so-called collaborative tagging is the process by which many users add metadata to shared content in the form of unstructured keywords. In their paper, they analyzed large sets of real-world data in order to explore and study social annotation and tagging with regard to their usefulness for web document classification. They were interested in finding out which kinds of documents are annotated more by end users than others, how users tend to annotate these documents, and, in particular, how these user-generated folksonomies are compared with the top-down taxonomies t...

Citation Context

...e communication. Some of them include § Flickr, ** Delicious, †† Bibsonomy, ‡‡ Technorati which are widely used portals for tagging. Basically tagging can be represented as Documents, Users and Tags. =-=[14]-=- This is sometimes known Collaborative Tagging. Since tags are used to describe the resources, to categorize the resources into a structured hierarchy tags play an important role for generating Taxono...

Citation Context

...ggest that tags seem to be better suited for classification of Web documents than anchor words or search keywords, whereas the last ones are more useful for information retrieval. In a further study (=-=Noll and Meinel, 2008a-=-), the same authors analyzed at which hierarchy depth tag-based classifiers can predict a category using DMOZ100k06 6 dataset with information from ODP and Del.icio.us. It was concluded that tags may ...