@inproceedings{shah-etal-2026-omniqum,
title = "{O}mni{QUM}: Multilingual Query-to-Location Understanding Model for Local Search",
author = "Shah, Avijit and
Dwarakanathan, Srini and
Kamdar, Yash and
Class, Danielle and
Moschitti, Alessandro",
editor = "Atanasova, Pepa and
Haffari, Gholamreza and
Wang, Minghan",
booktitle = "Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing: Industry Track",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.emnlp-industry.191/",
pages = "3217--3225",
ISBN = "979-8-89176-460-6",
abstract = "Local search systems must determine WHERE to search before retrieving results. We present OmniQUM, a fine-tuned 7B model for generative location disambiguation: given a query and geo-coordinates, it produces a disambiguated location target and bounding box decision in a single pass. Unlike NER + gazetteer pipelines, it jointly reasons over query intent and geographic context. We train across 14 locales and 7 languages with a failure-driven agentic pipeline generating targeted samples iteratively. We introduce expanding donut retrieval, sampling candidates from concentric, slightly overlapping rings to improve recall over naive bounding box retrieval. On 52K queries reflecting real traffic distribution, we achieve F1 0.94 on location disambiguation and 32.7{\%} retrieval recall improvement, surfacing 86K additional relevant POIs. Deployed with vLLM and bf16 on dual GPUs, we achieve sub-150ms TM95 latency at 0.12$\times$ baseline cost, with 7{--}38{\%} cache hit rates."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="shah-etal-2026-omniqum">
<titleInfo>
<title>OmniQUM: Multilingual Query-to-Location Understanding Model for Local Search</title>
</titleInfo>
<name type="personal">
<namePart type="given">Avijit</namePart>
<namePart type="family">Shah</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Srini</namePart>
<namePart type="family">Dwarakanathan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yash</namePart>
<namePart type="family">Kamdar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Danielle</namePart>
<namePart type="family">Class</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Alessandro</namePart>
<namePart type="family">Moschitti</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing: Industry Track</title>
</titleInfo>
<name type="personal">
<namePart type="given">Pepa</namePart>
<namePart type="family">Atanasova</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Gholamreza</namePart>
<namePart type="family">Haffari</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Minghan</namePart>
<namePart type="family">Wang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Budapest, Hungary</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-460-6</identifier>
</relatedItem>
<abstract>Local search systems must determine WHERE to search before retrieving results. We present OmniQUM, a fine-tuned 7B model for generative location disambiguation: given a query and geo-coordinates, it produces a disambiguated location target and bounding box decision in a single pass. Unlike NER + gazetteer pipelines, it jointly reasons over query intent and geographic context. We train across 14 locales and 7 languages with a failure-driven agentic pipeline generating targeted samples iteratively. We introduce expanding donut retrieval, sampling candidates from concentric, slightly overlapping rings to improve recall over naive bounding box retrieval. On 52K queries reflecting real traffic distribution, we achieve F1 0.94 on location disambiguation and 32.7% retrieval recall improvement, surfacing 86K additional relevant POIs. Deployed with vLLM and bf16 on dual GPUs, we achieve sub-150ms TM95 latency at 0.12\times baseline cost, with 7–38% cache hit rates.</abstract>
<identifier type="citekey">shah-etal-2026-omniqum</identifier>
<location>
<url>https://aclanthology.org/2026.emnlp-industry.191/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>3217</start>
<end>3225</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T OmniQUM: Multilingual Query-to-Location Understanding Model for Local Search
%A Shah, Avijit
%A Dwarakanathan, Srini
%A Kamdar, Yash
%A Class, Danielle
%A Moschitti, Alessandro
%Y Atanasova, Pepa
%Y Haffari, Gholamreza
%Y Wang, Minghan
%S Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing: Industry Track
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-460-6
%F shah-etal-2026-omniqum
%X Local search systems must determine WHERE to search before retrieving results. We present OmniQUM, a fine-tuned 7B model for generative location disambiguation: given a query and geo-coordinates, it produces a disambiguated location target and bounding box decision in a single pass. Unlike NER + gazetteer pipelines, it jointly reasons over query intent and geographic context. We train across 14 locales and 7 languages with a failure-driven agentic pipeline generating targeted samples iteratively. We introduce expanding donut retrieval, sampling candidates from concentric, slightly overlapping rings to improve recall over naive bounding box retrieval. On 52K queries reflecting real traffic distribution, we achieve F1 0.94 on location disambiguation and 32.7% retrieval recall improvement, surfacing 86K additional relevant POIs. Deployed with vLLM and bf16 on dual GPUs, we achieve sub-150ms TM95 latency at 0.12\times baseline cost, with 7–38% cache hit rates.
%U https://aclanthology.org/2026.emnlp-industry.191/
%P 3217-3225
Markdown (Informal)
[OmniQUM: Multilingual Query-to-Location Understanding Model for Local Search](https://aclanthology.org/2026.emnlp-industry.191/) (Shah et al., EMNLP 2026)
ACL