Deep Pairwise Learning To Rank For Search Autocomplete
This work addresses autocomplete ranking for commercial search engines, representing an incremental improvement over existing methods like LambdaMART.
The paper tackled the problem of improving autocomplete ranking in search engines by proposing DeepPLTR, a context-aware neural network pairwise ranker, which achieved a +3.90% MRR lift offline and a +0.06% GMV lift in an online A/B test.
Autocomplete (a.k.a "Query Auto-Completion", "AC") suggests full queries based on a prefix typed by customer. Autocomplete has been a core feature of commercial search engine. In this paper, we propose a novel context-aware neural network based pairwise ranker (DeepPLTR) to improve AC ranking, DeepPLTR leverages contextual and behavioral features to rank queries by minimizing a pairwise loss, based on a fully-connected neural network structure. Compared to LambdaMART ranker, DeepPLTR shows +3.90% MeanReciprocalRank (MRR) lift in offline evaluation, and yielded +0.06% (p < 0.1) Gross Merchandise Value (GMV) lift in an Amazon's online A/B experiment.