<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="ventirozos-etal-2026-useful">
    <titleInfo>
        <title>Useful Diversity or Hallucinated Variety? Persona-Conditioned Synthetic Query Generation for E-Commerce Search</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Filippos</namePart>
        <namePart type="family">Ventirozos</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Tom</namePart>
        <namePart type="family">Armitage</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Mahmoud</namePart>
        <namePart type="family">Oshagh</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Peter</namePart>
        <namePart type="given">A</namePart>
        <namePart type="family">Appleby</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Matthew</namePart>
        <namePart type="family">Shardlow</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2026-11</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <relatedItem type="host">
        <titleInfo>
            <title>Findings of the Association for Computational Linguistics: IJCNLP-AACL 2026</title>
        </titleInfo>
        <name type="personal">
            <namePart type="given">Yue</namePart>
            <namePart type="family">Zhang</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Ruifeng</namePart>
            <namePart type="family">Xu</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Ayu</namePart>
            <namePart type="family">Purwarianti</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Tong</namePart>
            <namePart type="family">Rong</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <name type="personal">
            <namePart type="given">Jun</namePart>
            <namePart type="family">Suzuki</namePart>
            <role>
                <roleTerm authority="marcrelator" type="text">editor</roleTerm>
            </role>
        </name>
        <originInfo>
            <publisher>Association for Computational Linguistics</publisher>
            <place>
                <placeTerm type="text">Hengqin, China</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">conference publication</genre>
        <identifier type="isbn">979-8-89176-444-6</identifier>
    </relatedItem>
    <abstract>Synthetic supervision can bootstrap semantic parsers in new search domains, but prompt-induced diversity is useful only when generations are faithful to the structured intent. We explore this for cold-start e-commerce search: 118,098 synthetic pairs produced by 6 LLMs with 9 persona-conditioned prompt families on a 47-field UK automotive schema, against 1,214 human-annotated and 15,304 production queries. We show that the IPIP personality bundle improves parser F1 in all four matched contrasts (Holm-corrected p \ensuremath&lt; .001), and all three bundles have a positive main effect over the complete 2⌃3 grid. Syntax conditioning is the axis that moves generations towards real traffic. It also induces lexical diversity. Alignment sKL, the divergence between von Mises-Fisher distributions fitted to synthetic and human-annotated queries, is an unlabelled proxy for parser F1 (rho = -0.917 on intent-elicited human queries, lower is closer). An LLM committee, checked against a blinded author pass, triages the 927 queries missed by every parser and labels 37% as hallucinated. Dropped constraints dominate. The four syntax-bearing families have the largest hallucinated shares, 38 to 42%, while IPIP alone has the lowest at 28%.</abstract>
    <identifier type="citekey">ventirozos-etal-2026-useful</identifier>
    <location>
        <url>https://aclanthology.org/2026.findings-aacl.20/</url>
    </location>
    <part>
        <date>2026-11</date>
        <extent unit="page">
            <start>350</start>
            <end>368</end>
        </extent>
    </part>
</mods>
</modsCollection>

%0 Conference Proceedings
%T Useful Diversity or Hallucinated Variety? Persona-Conditioned Synthetic Query Generation for E-Commerce Search
%A Ventirozos, Filippos
%A Armitage, Tom
%A Oshagh, Mahmoud
%A Appleby, Peter A.
%A Shardlow, Matthew
%Y Zhang, Yue
%Y Xu, Ruifeng
%Y Purwarianti, Ayu
%Y Rong, Tong
%Y Suzuki, Jun
%S Findings of the Association for Computational Linguistics: IJCNLP-AACL 2026
%D 2026
%8 November
%I Association for Computational Linguistics
%C Hengqin, China
%@ 979-8-89176-444-6
%F ventirozos-etal-2026-useful
%X Synthetic supervision can bootstrap semantic parsers in new search domains, but prompt-induced diversity is useful only when generations are faithful to the structured intent. We explore this for cold-start e-commerce search: 118,098 synthetic pairs produced by 6 LLMs with 9 persona-conditioned prompt families on a 47-field UK automotive schema, against 1,214 human-annotated and 15,304 production queries. We show that the IPIP personality bundle improves parser F1 in all four matched contrasts (Holm-corrected p \ensuremath< .001), and all three bundles have a positive main effect over the complete 2⌃3 grid. Syntax conditioning is the axis that moves generations towards real traffic. It also induces lexical diversity. Alignment sKL, the divergence between von Mises-Fisher distributions fitted to synthetic and human-annotated queries, is an unlabelled proxy for parser F1 (rho = -0.917 on intent-elicited human queries, lower is closer). An LLM committee, checked against a blinded author pass, triages the 927 queries missed by every parser and labels 37% as hallucinated. Dropped constraints dominate. The four syntax-bearing families have the largest hallucinated shares, 38 to 42%, while IPIP alone has the lowest at 28%.
%U https://aclanthology.org/2026.findings-aacl.20/
%P 350-368

Markdown (Informal)

[Useful Diversity or Hallucinated Variety? Persona-Conditioned Synthetic Query Generation for E-Commerce Search](https://aclanthology.org/2026.findings-aacl.20/) (Ventirozos et al., Findings 2026)

• Useful Diversity or Hallucinated Variety? Persona-Conditioned Synthetic Query Generation for E-Commerce Search (Ventirozos et al., Findings 2026)

ACL

• Filippos Ventirozos, Tom Armitage, Mahmoud Oshagh, Peter A. Appleby, and Matthew Shardlow. 2026. Useful Diversity or Hallucinated Variety? Persona-Conditioned Synthetic Query Generation for E-Commerce Search. In Findings of the Association for Computational Linguistics: IJCNLP-AACL 2026, pages 350–368, Hengqin, China. Association for Computational Linguistics.