Long Time No See: The Probability of Reusing Tags as a Function of Frequency and Recency
This work addresses tag recommendation for users in folksonomies by providing a more accurate and efficient model, though it is incremental as it builds on existing time-dependent mechanisms with a theory-driven twist.
The paper tackled tag recommendation by modeling human memory processes, using frequency and recency to estimate tag reuse probability, and showed that this approach outperforms conventional methods like 'most popular tags' and other established algorithms such as FolkRank and Collaborative Filtering on real-world datasets from BibSonomy, CiteULike, and Flickr.
In this paper, we introduce a tag recommendation algorithm that mimics the way humans draw on items in their long-term memory. This approach uses the frequency and recency of previous tag assignments to estimate the probability of reusing a particular tag. Using three real-world folksonomies gathered from bookmarks in BibSonomy, CiteULike and Flickr, we show how adding a time-dependent component outperforms conventional "most popular tags" approaches and another existing and very effective but less theory-driven, time-dependent recommendation mechanism. By combining our approach with a simple resource-specific frequency analysis, our algorithm outperforms other well-established algorithms, such as FolkRank, Pairwise Interaction Tensor Factorization and Collaborative Filtering. We conclude that our approach provides an accurate and computationally efficient model of a user's temporal tagging behavior. We show how effective principles for information retrieval can be designed and implemented if human memory processes are taken into account.