Long Time No See: The Probability of Reusing Tags As a Function of Frequency and Recency
D. Kowald, P. Seitlinger, C. Trattner, and T. Ley. Proceedings of the Companion Publication of the 23rd International Conference on World Wide Web Companion, page 463--468. Republic and Canton of Geneva, Switzerland, International World Wide Web Conferences Steering Committee, (2014)
DOI: 10.1145/2567948.2576934
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 incorporating 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 demonstrate how effective principles of information retrieval can be designed and implemented if human memory processes are taken into account.
%0 Conference Paper
%1 kowald2014probability
%A Kowald, Dominik
%A Seitlinger, Paul
%A Trattner, Christoph
%A Ley, Tobias
%B Proceedings of the Companion Publication of the 23rd International Conference on World Wide Web Companion
%C Republic and Canton of Geneva, Switzerland
%D 2014
%I International World Wide Web Conferences Steering Committee
%K aboutBibSonomy frequency recency reusing tagging
%P 463--468
%R 10.1145/2567948.2576934
%T Long Time No See: The Probability of Reusing Tags As a Function of Frequency and Recency
%U http://dx.doi.org/10.1145/2567948.2576934
%X 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 incorporating 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 demonstrate how effective principles of information retrieval can be designed and implemented if human memory processes are taken into account.
%@ 978-1-4503-2745-9
@inproceedings{kowald2014probability,
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 incorporating 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 demonstrate how effective principles of information retrieval can be designed and implemented if human memory processes are taken into account.},
acmid = {2576934},
added-at = {2014-10-01T17:39:40.000+0200},
address = {Republic and Canton of Geneva, Switzerland},
author = {Kowald, Dominik and Seitlinger, Paul and Trattner, Christoph and Ley, Tobias},
biburl = {https://www.bibsonomy.org/bibtex/2659fca43cb6751ee9b13297b797d91e1/sdo},
booktitle = {Proceedings of the Companion Publication of the 23rd International Conference on World Wide Web Companion},
description = {Long time no see},
doi = {10.1145/2567948.2576934},
interhash = {911a425fa4f6e883c32fa7a09840bdd8},
intrahash = {659fca43cb6751ee9b13297b797d91e1},
isbn = {978-1-4503-2745-9},
keywords = {aboutBibSonomy frequency recency reusing tagging},
location = {Seoul, Korea},
numpages = {6},
pages = {463--468},
publisher = {International World Wide Web Conferences Steering Committee},
series = {WWW Companion '14},
timestamp = {2014-10-01T17:39:40.000+0200},
title = {Long Time No See: The Probability of Reusing Tags As a Function of Frequency and Recency},
url = {http://dx.doi.org/10.1145/2567948.2576934},
year = 2014
}