analiza aiqbench
This work replaces fragmented, high-maintenance search pipelines with a unified Small Language Model that consolidates disparate query understanding functions into a single schema-constrained generator. By introducing Query Illuminator for auto-annotation and scalable evaluation, the approach effectively mitigates data bottlenecks while delivering improved engagement and lower costs on LinkedIn's infrastructure. The solution validates that consolidating heterogeneous tasks into one model can satisfy strict low-latency constraints even on limited GPU resources.
A Unified Structured Query Understanding Framework for Industrial Semantic Search
arXiv:2605.27441v2 Announce Type: replace-cross Abstract: Query understanding in large-scale industrial search systems is typically implemented as a cascade of disparate, task-specific components. While individually optimizable, this fragmented architecture incurs high maintenance overhead and results in inconsistent behaviors, particularly for long-tail queries. In this work, we propose and deploy a unified structured query understanding system that consolidates these heterogeneous functions into a single Small Language Model (SLM) that performs schema-constrained generation. To address the data bottlenecks inherent in unified modeling, we introduce Query Illuminator, a dual-purpose framework serving as: (i) a teacher model for high-quality auto-annotation and distillation, and (ii) a surrogate judge for scalable evaluation where human labels are scarce. We validate this approach through extensive offline and online tests within LinkedIn's Job Search system. Furthermore, we demonstrate the framework's horizontal extensibility through a cross-domain case study on People Search. The results show improved user engagement and reduced operational costs, achieved while satisfying strict low-latency serving constraints on limited GPU resources.
Abstract:Query understanding in large-scale industrial search systems is typically implemented as a cascade of disparate, task-specific components. While individually optimizable, this fragmented architecture incurs high maintenance overhead and results in inconsistent behaviors, particularly for long-tail queries. In this work, we propose and deploy a unified structured query understanding system that consolidates these heterogeneous functions into a single Small Language Model (SLM) that performs schema-constrained generation. To address the data bottlenecks inherent in unified modeling, we introduce Query Illuminator, a dual-purpose framework serving as: (i) a teacher model for high-quality auto-annotation and distillation, and (ii) a surrogate judge for scalable evaluation where human labels are scarce. We validate this approach through extensive offline and online tests within LinkedIn's Job Search system. Furthermore, we demonstrate the framework's horizontal extensibility through a cross-domain case study on People Search. The results show improved user engagement and reduced operational costs, achieved while satisfying strict low-latency serving constraints on limited GPU resources.
Information Retrieval (cs.IR); Machine Learning (cs.LG)
arXiv:2605.27441 [cs.IR]
(or arXiv:2605.27441v2 [cs.IR] for this version)
https://doi.org/10.1145/3770855.3818312 DOI(s) linking to related resources
Text integral sub licența deschisă a publicației · 176 cuvinte
Subiecte conexe
Modernizing and scaling support operations with generative AI on AWS
AWS · 2 sept. 2026
Manage agents, tools and skills at scale with AWS Agent Registry
AWS · 31 aug. 2026
Presentation: Designing AI Platforms for Reliability: Tools for Certainty, Agents for Discovery
InfoQ · 27 mai 2026
What to expect at Dell Technologies New York Symposium: Join theCUBE Oct. 20
SiliconANGLE · 16 oct. 2026