Building recommenders without machine learning:
Where the privacy risks of recommender systems live:
The gap between published research and production:
People get excited about the wrong contextual features:
Recommending more similar items can be good, bad, and everything in between:
A common misunderstanding about what recommender systems "know":
The difficulty of applying bug bounties to attacks against recommender systems:
Recommender systems aren't a tool for analyzing your data:
Recommender systems are a tool for scaling human expertise:
Privacy-preserving recommenders are hard:
A hack for coping with sparsity in recommender systems:
Recommender systems don't know when they're spewing out nonsensical recs:
The organizational dynamics and trade-offs that crop up when rolling out algorithmic home screen ranking:
How to organize product teams for recommendation-based products:
