The impact of ambiguity and redundancy on tag recommendation in folksonomies

Abstract

Collaborative tagging applications have become a popular tool allowing Internet users to manage online resources with tags. Most collaborative tagging applications permit unsupervised tagging resulting in tag ambiguity in which a single tag has many different meanings and tag redundancy in which several tags have the same meaning. Common metrics for evaluating tag recommenders may overestimate the utility of ambiguous tags or ignore the appropriateness of redundant tags. Ambiguity and redundancy may even burden the user with additional effort by requiring them to clarify an annotation or forcing them to distinguish between highly related items. In this paper we demonstrate that ambiguity and redundancy impede the evaluation and performance of tag recommenders. Five tag recommendation strategies based on popularity, collaborative filtering and link analysis are explored. We use a cluster-based approach to define ambiguity and redundancy and provide extensive evaluation on three real world datasets.

Semantic Disambiguation and Contextualisation of Social Tags

“…As shown in previous works [1,12,16,18], semantic disambiguation and contextualisation of social tags can be used to improve folksonomy-based personalised search and recommendation strategies. Recently, in [5], we have preliminary evaluated our approach with a number of state of the art recommenders [6] on a Delicious dataset, and have obtained 13% to 24% precision/recall improvements by only contextualising 5.3% of the tags available in that dataset.…”

“…As representative examples, as for October 2011, Wikipedia contains over 200K disambiguation entries 8 , and Gemmell et al [12] demonstrate that ambiguity and redundancy impede the evaluation and performance of tag-based recommender systems, especially in folksonomies that include broad domains.…”

Semantic Contextualisation of Social Tag-Based Profiles and Item Recommendations

“…These approaches usually involve the application of clustering techniques over the co-occurrence information gathered from the folksonomy [3,4,20], and have been exploited by recent personalisation and recommendation approaches [8,17]. Their main advantage is that an external knowledge source is not required.…”

“…Current approaches have to define a stop criterion for the clustering processes. For instance, a hierarchical clustering [17] needs to establish the proper level at which clusters are selected, whereas an approach using a partitional clustering technique such as K-means needs to define beforehand how many clusters to build [8]. These values are difficult to define without proper evaluation, and have a definite impact on the outcome of the clustering process, and ultimately, on the semantic disambiguation or contextualisation approach.…”

How annotation styles influence content and preferences

“…Tagging, or more generally, annotation of digital objects has been an area of interest in recent years, with research done primarily in the whys, whats and hows of tagging, including the analysis of tagging vocabularies, indexing, and recommender systems for tags [22,8,7]. Building on existing understandings of annotation behavior, this paper focuses instead on annotation system design and its impact on users.…”