An LLM judge uses semantic matching on user text to deliver reliable, explainable top-K recommendation scores instead of biased ID-based holdout metrics.
arXiv:2504.03997
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.IR 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
LLM-as-a-Judge for Reliable and Explainable Offline Evaluation in Top-K Recommendation
An LLM judge uses semantic matching on user text to deliver reliable, explainable top-K recommendation scores instead of biased ID-based holdout metrics.