Welcome to the InfoSci Platform
Reference Hub1
Towards an e-Science Environment for Collaborative Filtering Researchers

Towards an e-Science Environment for Collaborative Filtering Researchers

Nikos Manouselis, Giannis Stoitsis
Copyright: © 2014 |Volume: 4 |Issue: 1 |Pages: 32
ISSN: 1947-9077|EISSN: 1947-9085|EISBN13: 9781466653597|DOI: 10.4018/ijdls.2014010104
Cite Article Cite Article

MLA

Manouselis, Nikos, and Giannis Stoitsis. "Towards an e-Science Environment for Collaborative Filtering Researchers." IJDLS vol.4, no.1 2014: pp.41-72. http://doi.org/10.4018/ijdls.2014010104

APA

Manouselis, N. & Stoitsis, G. (2014). Towards an e-Science Environment for Collaborative Filtering Researchers. International Journal of Digital Library Systems (IJDLS), 4(1), 41-72. http://doi.org/10.4018/ijdls.2014010104

Chicago

Manouselis, Nikos, and Giannis Stoitsis. "Towards an e-Science Environment for Collaborative Filtering Researchers," International Journal of Digital Library Systems (IJDLS) 4, no.1: 41-72. http://doi.org/10.4018/ijdls.2014010104

Export Reference

Mendeley
Favorite Full-Issue Download

Abstract

A plethora of collaborative filtering algorithms have been proposed in related literature. Due to the dynamic and changing parameters of the various application contexts, careful testing and parameterization has to be carried out before an algorithm is finally deployed in a real setting. This paper investigates how a previously proposed tool for simulated testing of collaborative filtering algorithms, called the Collaborative Filtering Simulator (CollaFiS), can be expanded to an e-science environment for researchers so that it runs over a digital research infrastructure. More specifically, a survey of design options for neighborhood-based collaborative filtering systems is carried out in order to illustrate the variety of requirements that need to be met. Then, a number of usage scenarios that could be supported by an e-science environment for collaborative filtering research are presented. Three example dataset cases are used to illustrate how the new version of CollaFiS can support research and experimentation on collaborative filtering algorithms using different data and various parameter options. Overall, this paper showcases how e-science environments and infrastructures may facilitate the research activities of people working on recommender systems.

Request Access

You do not own this content. Please login to recommend this title to your institution's librarian or purchase it from the IGI Global bookstore.