Long Time No See: The Probability of Reusing Tags as a Function of Frequency and Recency

TL;DR

This paper presents a tag recommendation algorithm based on human memory principles, utilizing frequency and recency of tags to improve accuracy over traditional methods across multiple real-world datasets.

Contribution

Introduces a novel time-dependent tag recommendation algorithm inspired by human memory, outperforming existing methods in real-world folksonomies.

Findings

Outperforms conventional 'most popular tags' approaches

Surpasses existing time-dependent recommendation mechanisms

Achieves higher accuracy and efficiency in modeling user tagging behavior

Abstract

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.