RADAR precomputes top-200 lists per user offline with a full ranking model over a 50X candidate pool, reporting 16.5% recall@200 versus 8.1% for DNN and a +0.8% engagement lift.
This is the standard item-based collaborative filtering method 3
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
baseline 1
citation-polarity summary
fields
cs.IR 1years
2025 1verdicts
CONDITIONAL 1roles
baseline 1polarities
baseline 1representative citing papers
citing papers explorer
-
RADAR: Recall Augmentation through Deferred Asynchronous Retrieval
RADAR precomputes top-200 lists per user offline with a full ranking model over a 50X candidate pool, reporting 16.5% recall@200 versus 8.1% for DNN and a +0.8% engagement lift.