Djoerd Hiemstra πIgor Jakovljevic presents CERN Recommendations: a recommender system using privacy-preserving collaborative filtering. #ossym2023
Maximilian Theiner and Patrick Lovric present a search system to find repair shops and cafes for digital and other tools, based on your geographic location using @openstreetmap. #ossym2023
Reihaneh Manteghi presents a customizable literature search system that uses the categorization of documents. #ossym2023
Viktor Lofgren of Marginala asks today's most important question:
Why does the web look like a shopping maill?
(through the eyes of search engines, social media platforms, forums, ...)
Not on Marginalia: https://search.marginalia.nu/
Marginalia Search
search.marginalia.nu is a small independent do-it-yourself search engine for surprising but content-rich websites that never ask you to accept cookies or subscribe to newsletters. The goal is to bring you the sort of grass fed, free range HTML your grandma used to write.
Anke Meinders from https://fragfinn.de discusses one of my favorite (and important) topics: Search for Children. #ossym2023
Colin Hayhurst gives a excellent insight into the nuts and bolts of the @Mojeek search engine. #ossym2023
About popularity bias: We have to explore more (and maybe earn less) to serve the long tail of people. Like nature being strong by a diverse ecosystem, the information ecosystem needs a long tail over different results. #ossym2023
Degenerate feedback loops in recommender systems by Jiang et al. discusses echo chambers that race to the bottom. #ossym2023
Engagement bias: What percentage produces 50% of the content? #ossym2023
7% of Facebook
4% on Amazon
2% on Twitter
0.04% on Wikipedia
Wisdom of Crowds or Wisdom of a Few? by Ricardo Baeza-Yates et al.
The famous Ricardo Baeza-Yates talks about bias in search. #ossym2023