Paper grounds recommender similarity measures in probability, reframing cosine similarity
_reachsumit · x · 2026-10-28
Paper (arXiv:2610.34402) gives a principled approach to designing item similarity measures for neighborhood-based collaborative filtering:
- Motivation: many common similarity measures are unjustified heuristics, forcing empirical trial-and-error.
- Method: grounded in probabilistic modeling and parameter estimation, the authors derive two previously overlooked measures from Bernoulli and multinomial Naive Bayes.
- Ablations: highlight the critical roles of bias, probabilistic smoothing, handling unknowns correctly, and distribution choice.
- Insight: cosine similarity is shown to be the solution to a specific optimization problem, reframing this decades-old heuristic.
More from Research
- TTME beats test-time training on 5 math tasks while cutting FLOPs by 82.8-84.6% — UCSantaBarbara · 2026-10-28
- EnvLunch turns executable environments into free verifiable agent training data — KingsCollegeLondon · 2026-10-28
- Anthropic reveals it is developing its own experimental drugs using Claude — Polymarket · 2026-10-28
- Stanford launches TaskNet: 132 worker interviews build a large-scale occupational task inventory — soumitrashukla9 · 2026-10-28
- KAIST's Principia uses LLMs to construct and refine scientific principles at test time — coallaoh · 2026-10-28
- The AlphaFold Data Problem: Quantum Chemistry Models as Training Data for Larger Scales — TimothyDuignan · 2026-10-28