Search papers, labs, and topics across Lattice.
This paper introduces IntentTune, a framework designed to resolve ambiguous user queries in e-commerce search by leveraging user-specific behavioral signals and population-level demand patterns. The research highlights that while population-level data is inadequate for accurately inferring user intent from under-specified queries, individual user behaviors, particularly prior search queries, significantly enhance intent detection. The key finding reveals that incorporating personalized signals leads to superior performance in identifying attributes such as gender, age group, and product category, thereby improving the relevance of search results.
User-specific signals can dramatically enhance intent detection accuracy in e-commerce search, outperforming traditional population-level approaches.
Understanding user intent is fundamental to delivering relevant search results in e-commerce. However, substantial fraction of real-world queries are under-specified (e.g., "watch" or "shirt"), lacking explicit attributes such as gender or age group. This ambiguity poses a significant challenge for query intent detection models in e-commerce search systems, which must accurately infer latent user intent (e.g., age, gender) to support effective downstream retrieval. We introduce IntentTune, a framework for resolving ambiguous or under-specified query intents by leveraging either (1) user-specific behavioral signals including search history, browsing activity, and profile attributes or (2) population-level demand patterns aggregated across all users. Through experiments on real-world e-commerce data, we first demonstrate that population-level demand patterns alone are insufficient to reliably infer intent in under-specified queries. We then demonstrate that user-specific behavioral signals -- particularly prior search queries -- outperform both population-level statistics and static profile information for inferring gender, age group, product category, and size intent from underspecified queries.
