Abstract
When someone asks where machine learning is broadly applied in an industrial context, recommender systems is a typical answer. Indeed, these systems are ubiquitous, and we rely on them a lot. Amazon is maybe the best example of an e-commerce site that utilizes many types of recommendations to enhance users’ experience and help them quickly find what they are looking for. Spotify, whose domain is music, is another good example. Despite heavy usage of machine learning, recommender systems differ in two crucial ways from classically trained ones:
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
MovieLens is part of the GroupLens research project (see https://grouplens.org ), which allows you to download free datasets composed of anonymized movie ratings. We will use some of this data in subsequent examples (see reference [1]). Another associated project is LensKit (see https://lenskit.org ) for building experimental recommender systems in Python (some earlier editions were Java based). You will see examples using this framework.
- 2.
At the time of this writing, the stable version is 0.5.0, which doesn’t yet contain the ndcg function (the latest development version already has it). So, you may expect the metrics API to change in the future in incompatible ways.
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 2019 Ervin Varga
About this chapter
Cite this chapter
Varga, E. (2019). Recommender Systems. In: Practical Data Science with Python 3. Apress, Berkeley, CA. https://doi.org/10.1007/978-1-4842-4859-1_8
Download citation
DOI: https://doi.org/10.1007/978-1-4842-4859-1_8
Published:
Publisher Name: Apress, Berkeley, CA
Print ISBN: 978-1-4842-4858-4
Online ISBN: 978-1-4842-4859-1
eBook Packages: Professional and Applied ComputingApress Access BooksProfessional and Applied Computing (R0)