UAI 2026

We propose the DenoiseBid method, which corrects the generated CTRs and CVRs to make the resulting bids more efficient in auctions.

KDD 2026

CTR models rank new items poorly until they gather enough clicks. UCB-FE wraps any CTR model with position-aware upper confidence bounds at inference time, boosting cold items — no retraining, no architecture changes — and comes with a new coldNDCG metric to measure it.

SIGIR 2026

Search and recommender systems have produced highly relevant search results. A natural next step in the development of such systems in e-commerce is to rerank these results to increase the platform's revenue from paid promotion products.

SIGIR 2026

We introduce RRE-GTC (Ru-Real-Estate Geo-Temporal Clusters), an open resource designed to benchmark ranking, pricing, and recommendation algorithms in dynamic spatial contexts. Derived from a massive collection of apartment listings in major Russian cities, RRE-GTC aggregates 456,182 geo-temporal clusters, offering a privacy-preserving alternative to releasing raw listing-level data.

The Bulletin of Irkutsk State University. Series Mathematics

We introduce LLM-based Plan Mapping, a framework that embeds the default execution plan of a query, finds its k nearest neighbors among previously executed plans, and recommends database hintsets based on neighborhood voting.