Print Email Facebook Twitter Towards Minimal Necessary Data Title Towards Minimal Necessary Data: The Case for Analyzing Training Data Requirements of Recommender Algorithms Author Larson, M.A. (TU Delft Multimedia Computing; Radboud Universiteit Nijmegen) Zito, Alessandro (Politecnico di Milano) Loni, B. (TU Delft Multimedia Computing) Cremonesi, Paolo (Politecnico di Milano) Date 2017 Abstract This paper states the case for the principle of minimal necessary data: If two recommender algorithms achieve the same effectiveness, the better algorithm is the one that requires less user data. Applying this principle involves carrying out training data requirements analysis, which we argue should be adopted as best practice for the development and evaluation of recommender algorithms. We takethe position that responsible recommendation is recommendation that serves the people whose data it uses. To minimize the imposition on users’ privacy, it is important that a recommender system does not collect or store more user information than it absolutely needs. Further, algorithms using minimal necessary data reduce training time and address the cold start problem. To illustrate the trade-off between training data volume and accuracy, we carry outa set of classic recommender system experiments. We conclude thatconsistently applying training data requirements analysis would represent a relatively small change in researchers’ current practices, but a large step towards more responsible recommender systems. To reference this document use: http://resolver.tudelft.nl/uuid:03b3462d-ef44-465b-b02f-011dc7a1574d DOI https://doi.org/10.18122/B2VX12 Source FATREC Workshop on Responsible Recommendation Proceedings Event FATREC 2017, 2017-08-31, Como, Italy Part of collection Institutional Repository Document type conference paper Rights © 2017 M.A. Larson, Alessandro Zito, B. Loni, Paolo Cremonesi Files PDF 35744357.pdf 785.62 KB Close viewer /islandora/object/uuid:03b3462d-ef44-465b-b02f-011dc7a1574d/datastream/OBJ/view