Harnessing the power of social bookmarking for improving tag-based recommendations

摘要: Used the collaborative tagging idea to produce personalized recommendations.Attempted utilization of power taxonomies through clustering annotations.Experimental evaluation model using data from public annotation system, citeUlike.The approach improved prediction quality without compromising computational efficiency. Social bookmarking and has emerged a new era in user collaboration. Collaborative Tagging allows users annotate content their liking, which via appropriate algorithms can render useful for provision product recommendations. It is case today tag-based work complementary rating-based recommendation mechanisms predict liking various products. In this paper we propose an alternative algorithm computing recommendations products, that uses exclusively tags provided by users. Our based on semantic similarity user-provided them into groups similar meaning. Afterwards, some measurable characteristics users' Annotation Competency are combined with other metrics, such as similarity, predictions. The used real-world citeUlike, confirmed our outperforms baseline Vector Space model, well state art algorithms, predicting more accurately.