Building an Open-Source Online Learn-to-Rank Engine - Haystack EU 2022

Discover how to build an open-source online Learn-to-rank engine in this 43-minute conference talk from Haystack EU 2022. Explore the challenges of implementing personalized search relevancy and learn about Metarank, an open-source personalization service designed to handle common data and feature engineering tasks for Learn-to-rank (LTR) applications. Gain insights into leveraging visitor behavior for LTR tasks, implementing advanced features like sliding window counters, per-item conversion and CTR rates, and customer profile tracking. Understand how Metarank simplifies the process of reordering items in real-time to maximize goals such as Click-Through Rate (CTR) using only a YAML config and JSON I/O. Presented by Roman Grebennikov, an independent search engineer specializing in relevancy, personalization, and recommendations, this talk offers practical knowledge for those interested in open-source software, functional programming, learn-to-rank models, and performance tuning in search applications.

Haystack EU 2022 - Roman Grebennikov: Building an open-source online Learn-to-rank engine