HierCat: Hierarchical Query Categorization from Weakly Supervised Data at Facebook Marketplace
This work addresses the problem of improving search relevance for users on e-commerce platforms like Facebook Marketplace, though it is incremental as it builds on existing dual-encoder architectures.
The paper tackles query categorization at Facebook Marketplace by developing HierCat, a system that uses multi-task pre-training and hierarchical inference on weakly supervised data, resulting in a 1.4% improvement in NDCG and a 4.3% increase in searcher engagement in online tests.
Query categorization at customer-to-customer e-commerce platforms like Facebook Marketplace is challenging due to the vagueness of search intent, noise in real-world data, and imbalanced training data across languages. Its deployment also needs to consider challenges in scalability and downstream integration in order to translate modeling advances into better search result relevance. In this paper we present HierCat, the query categorization system at Facebook Marketplace. HierCat addresses these challenges by leveraging multi-task pre-training of dual-encoder architectures with a hierarchical inference step to effectively learn from weakly supervised training data mined from searcher engagement. We show that HierCat not only outperforms popular methods in offline experiments, but also leads to 1.4% improvement in NDCG and 4.3% increase in searcher engagement at Facebook Marketplace Search in online A/B testing.