Conference Paper

Exploring the Influence of Tagging Motivation on Tagging Behavior

Authors:
To read the full-text of this research, you can request a copy directly from the authors.

Abstract

The reasons why users tag have remained mostly elusive to quantitative investigations. In this paper, we distinguish between two types of motivation for tagging: While categorizers use tags mainly for categorizing resources for later browsing, describers use tags mainly for describing resources for later retrieval. To characterize users with regard to these different motivations, we introduce statistical measures and apply them to 7 different real-world tagging datasets. We show that while most taggers use tags for both categorizing and describing resources, different tagging systems lend themselves to different motivations for tagging. Additionally we show that the distinction between describers and categorizers can improve the performance of tag recommendation.

No full-text available

Request Full-text Paper PDF

To read the full-text of this research,
you can request a copy directly from the authors.

... The test set was used to measure the efficiency of the recommendation systems. Similar partitioning schemes (90% training, 10% test) are used by other researchers (Kern et al., 2010;. We had hidden the tags of the test resources, as they were used for evaluation purposes. ...
Article
Full-text available
Purpose Tags are used to annotate resources on social media platforms. Most tag recommendation methods use popular tags, but in the case of new resources that are as yet untagged (the cold start problem), popularity-based tag recommendation methods fail to work. The purpose of this paper is to propose a novel model for tag recommendation called multi-feature space latent Dirichlet allocation (MFS-LDA) for cold start problem. Design/methodology/approach MFS-LDA is a novel latent Dirichlet allocation (LDA)-based model which exploits multiple feature spaces (title, contents, and tags) for recommending tags. Exploiting multiple feature spaces allows MFS-LDA to recommend tags even if data from a feature space is missing (the cold start problem). Findings Evaluation of a publicly available data set consisting of around 20,000 Wikipedia articles that are tagged on a social bookmarking website shows a significant improvement over existing LDA-based tag recommendation methods. Originality/value The originality of MFS-LDA lies in segregation of features for removing bias toward dominant features and in synchronization of multiple feature space for tag recommendation.
... While we believe that concept recommender systems could represent a useful instrument to augment user experience on collaborative ontology engineering platforms in the biomedical domain, we have not performed evaluations of our implementations yet. Understanding what kinds of recommender techniques are useful in what kind of contexts, including a more sophisticated analysis of user-behavior similar to (Kern et al., 2010), represent important next steps for future research. We believe that understanding the utility of recommender systems to steer and augment user activity in collaborative ontology engineering projects represents an exciting avenue for future biomedical research. ...
Conference Paper
Full-text available
Biomedical ontologies such as the 11th revision of the International Classification of Diseases and others are increasingly produced with the help of collaborative ontology engineering platforms that facilitate cooperation and coordination among a large number of users and contributors. While collaborative approaches to engineering biomedical ontologies can be expected to yield a number of advantages, such as increased participation and coverage, they come with a number of novel challenges and risks. For example, they might suffer from low participation, lack of coordination, lack of control or other related problems that are neither well understood nor addressed by the current state of research. In this paper, we aim to tackle some of these problems by exploring techniques for recommending concepts to experts on collaborative ontology engineering platforms. In detail, this paper will (i) discuss different recommendation techniques from the literature (ii) map and apply these categories to the domain of collaboratively engineered biomedical ontologies and (iii) present prototypical implementations of selected recommendation techniques as a proof-of-concept.
Conference Paper
Full-text available
Social tagging systems allow users to upload and assign keywords to digital resources. Thus a body of user annotated resources gradually evolves: Users can share resources, re-find their own resources or use the systems as search engines for items added by the whole user population. In this paper we want to contribute towards a better understanding of usage patterns within social tagging systems by presenting results from a survey of 142 users of the systems Flickr, Youtube, Delicious and Connotea. Data was gathered partly by using the Mechanical Turk service, and partly via an announcement on the Connotea blog. Our study reveals differences of user motivation and tag usage between systems. While (resource) sharing emerges as an all-embracing intra-system motivation, users differ with respect to social spheres of sharing. Based on our results which we integrated with earlier research from Cool and Belkin (2002), we propose a model of information behaviour in social tagging systems.
Conference Paper
While recent research has advanced our understanding about the structure and dynamics of social tagging systems, we know little about (i) the underlying motivations for tagging (why users tag), and (ii) how they influence the properties of resulting tags and folksonomies. In this paper, we focus on problem (i) based on a distinction between two types of user motivations that we have identified in earlier work: Categorizers vs. Describers. To that end, we systematically define and evaluate a number of measures designed to discriminate between describers, i.e. users who use tags for describing resources as opposed to categorizers, i.e. users who use tags for categorizing resources. Subsequently, we present empirical findings from qualitative and quantitative evaluations of the measures on real world tagging behavior. In addition, we conducted a recommender evaluation in which we study the effectiveness of each of the presented measures and found the measure based on the tag content to be the most accurate in predicting the user behavior closely followed by a content independent measure. The overall contribution of this paper is the presentation of empirical evidence that tagging motivation can be approximated with simple statistical measures. Our research is relevant for (a) designers of tagging systems aiming to better understand the motivations of their users and (b) researchers interested in studying the effects of users' tagging motivation on the properties of resulting tags and emergent structures in social tagging systems
Conference Paper
Recent research provides evidence for the presence of emergent semantics in collaborative tagging systems. While several methods have been proposed, little is known about the factors that influence the evolution of semantic structures in these systems. A natural hypothesis is that the quality of the emergent semantics depends on the pragmatics of tagging: Users with certain usage patterns might contribute more to the resulting semantics than others. In this work, we propose several measures which enable a pragmatic differentiation of taggers by their degree of contribution to emerging semantic structures. We distinguish between categorizers, who typically use a small set of tags as a replacement for hierarchical classification schemes, and describers, who are annotating resources with a wealth of freely associated, descriptive keywords. To study our hypothesis, we apply semantic similarity measures to 64 different partitions of a real-world and large-scale folksonomy containing different ratios of categorizers and describers. Our results not only show that "verbose" taggers are most useful for the emergence of tag semantics, but also that a subset containing only 40% of the most 'verbose' taggers can produce results that match and even outperform the semantic precision obtained from the whole dataset. Moreover, the results suggest that there exists a causal link between the pragmatics of tagging and resulting emergent semantics. This work is relevant for designers and analysts of tagging systems interested (i) in fostering the semantic development of their platforms, (ii) in identifying users introducing "semantic noise", and (iii) in learning ontologies.