@inproceedings{kuissi-etal-2026-retrieval,
title = "Retrieval Evaluation Under Temporal Drift in Technical Documentation",
author = "Kuissi, Nathan and
Subrahmanyan, Suraj and
Thakur, Nandan and
Lin, Jimmy",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {EMNLP} 2026",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-emnlp.1754/",
pages = "37517--37527",
ISBN = "979-8-89176-462-0",
abstract = "Information retrieval (IR) benchmarks typically follow the Cranfield paradigm, relying on static and predefined corpora. However, temporal changes in technical corpora, such as API deprecations and code reorganizations, can render existing benchmarks stale. In our work, we investigate how temporal corpus drift affects FreshStack, a retrieval benchmark focused on technical domains. We examine two independent corpus snapshots of FreshStack from October 2024 and October 2025 to answer queries about LangChain. Our analysis shows that all but one query posed in 2024 remains fully supported by the 2025 corpus, as relevant documents *migrate* from LangChain to competitor repositories, such as LlamaIndex. Next, we compare the accuracy of retrieval models on both snapshots and observe only minor shifts in model rankings, with overall strong correlation of up to 0.978 Kendall $\tau$ at Recall@50. These results suggest that retrieval benchmarks for technical documentation, e.g., LangChain, can remain reliable when re-judged in the drift regime we observe, where relevant evidence is *relocated* rather than invalidated. All resources (code and data) are publicly available at https://github.com/fresh-stack/driftbench."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="kuissi-etal-2026-retrieval">
<titleInfo>
<title>Retrieval Evaluation Under Temporal Drift in Technical Documentation</title>
</titleInfo>
<name type="personal">
<namePart type="given">Nathan</namePart>
<namePart type="family">Kuissi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Suraj</namePart>
<namePart type="family">Subrahmanyan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Nandan</namePart>
<namePart type="family">Thakur</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jimmy</namePart>
<namePart type="family">Lin</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>Findings of the Association for Computational Linguistics: EMNLP 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hung-Yi</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sunipa</namePart>
<namePart type="family">Dev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jessy</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Desmond</namePart>
<namePart type="family">Elliott</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">979-8-89176-462-0</identifier>
</relatedItem>
<abstract>Information retrieval (IR) benchmarks typically follow the Cranfield paradigm, relying on static and predefined corpora. However, temporal changes in technical corpora, such as API deprecations and code reorganizations, can render existing benchmarks stale. In our work, we investigate how temporal corpus drift affects FreshStack, a retrieval benchmark focused on technical domains. We examine two independent corpus snapshots of FreshStack from October 2024 and October 2025 to answer queries about LangChain. Our analysis shows that all but one query posed in 2024 remains fully supported by the 2025 corpus, as relevant documents *migrate* from LangChain to competitor repositories, such as LlamaIndex. Next, we compare the accuracy of retrieval models on both snapshots and observe only minor shifts in model rankings, with overall strong correlation of up to 0.978 Kendall τ at Recall@50. These results suggest that retrieval benchmarks for technical documentation, e.g., LangChain, can remain reliable when re-judged in the drift regime we observe, where relevant evidence is *relocated* rather than invalidated. All resources (code and data) are publicly available at https://github.com/fresh-stack/driftbench.</abstract>
<identifier type="citekey">kuissi-etal-2026-retrieval</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.1754/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>37517</start>
<end>37527</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Retrieval Evaluation Under Temporal Drift in Technical Documentation
%A Kuissi, Nathan
%A Subrahmanyan, Suraj
%A Thakur, Nandan
%A Lin, Jimmy
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Findings of the Association for Computational Linguistics: EMNLP 2026
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-462-0
%F kuissi-etal-2026-retrieval
%X Information retrieval (IR) benchmarks typically follow the Cranfield paradigm, relying on static and predefined corpora. However, temporal changes in technical corpora, such as API deprecations and code reorganizations, can render existing benchmarks stale. In our work, we investigate how temporal corpus drift affects FreshStack, a retrieval benchmark focused on technical domains. We examine two independent corpus snapshots of FreshStack from October 2024 and October 2025 to answer queries about LangChain. Our analysis shows that all but one query posed in 2024 remains fully supported by the 2025 corpus, as relevant documents *migrate* from LangChain to competitor repositories, such as LlamaIndex. Next, we compare the accuracy of retrieval models on both snapshots and observe only minor shifts in model rankings, with overall strong correlation of up to 0.978 Kendall τ at Recall@50. These results suggest that retrieval benchmarks for technical documentation, e.g., LangChain, can remain reliable when re-judged in the drift regime we observe, where relevant evidence is *relocated* rather than invalidated. All resources (code and data) are publicly available at https://github.com/fresh-stack/driftbench.
%U https://aclanthology.org/2026.findings-emnlp.1754/
%P 37517-37527
Markdown (Informal)
[Retrieval Evaluation Under Temporal Drift in Technical Documentation](https://aclanthology.org/2026.findings-emnlp.1754/) (Kuissi et al., Findings 2026)
ACL