The effect of the Digital Services Act (DSA) on recommender systems in the EU
Luca Bertuzzi at Euractiv writes about the Digital Services Act (DSA), and how this would affect recommender systems.
[…]Schaldemose wanted to turn off online platforms’ recommender systems by default. As a result, social media like Facebook or Instagram would not personalise the user experience based on individual preferences unless people opted into that option.
While the proposal did not fly, Article 29 was revised to commend very large online platforms to provide at least one option that is not based on profiling. In addition, users should be able “to select and to modify at any time their preferred option for each of the recommender systems that determines the relative order of information presented to them.”
Furthermore, all online platforms will need to make their algorithm’s ‘main parameters’ to rank content and personalise the user experience easily accessible. That includes, as a minimum, the list of criteria used to assess content relevance, the objective of the optimisation, how the system interacts with user behaviour, and the relative weight of each parameter.
The Impact of the Digital Services Act (DSA) on Recommender Systems [Mozilla]
Big tech’s ‘blackbox’ [recommendation] algorithms face regulatory oversight under EU plans [Natasha Lomas @ TechCrunch]
Should Recommender Systems Be Exempt From the Post-Tracking Age? [Martin Anderson@Unite.ai
About The Author
Joeran Beel
I am the founder of Recommender-Systems.com and head of the Intelligent Systems Group (ISG) at the University of Siegen, Germany https://isg.beel.org. We conduct research in recommender-systems (RecSys), personalization and information retrieval (IR) as well as on automated machine learning (AutoML), meta-learning and algorithm selection. Domains we are particularly interested in include smart places, eHealth, manufacturing (industry 4.0), mobility, visual computing, and digital libraries. We founded or maintain, among others, LensKit-Auto, Darwin & Goliath, Mr. DLib, and Docear, each with thousand of users; we contributed to TensorFlow, JabRef and others; and we developed the first prototypes of automated recommender systems (AutoSurprise and Auto-CaseRec) and Federated Meta Learning (FMLearn Server and Client).