<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="kutuzov-etal-2022-contextualized">
    <titleInfo>
        <title>Contextualized embeddings for semantic change detection: Lessons learned</title>
    </titleInfo>
    <name type="personal">
        <namePart type="given">Andrey</namePart>
        <namePart type="family">Kutuzov</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Erik</namePart>
        <namePart type="family">Velldal</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <name type="personal">
        <namePart type="given">Lilja</namePart>
        <namePart type="family">Øvrelid</namePart>
        <role>
            <roleTerm authority="marcrelator" type="text">author</roleTerm>
        </role>
    </name>
    <originInfo>
        <dateIssued>2022</dateIssued>
    </originInfo>
    <typeOfResource>text</typeOfResource>
    <genre authority="bibutilsgt">journal article</genre>
    <relatedItem type="host">
        <titleInfo>
            <title>Northern European Journal of Language Technology</title>
        </titleInfo>
        <originInfo>
            <issuance>continuing</issuance>
            <publisher>Linköping University Electronic Press</publisher>
            <place>
                <placeTerm type="text">Linköping, Sweden</placeTerm>
            </place>
        </originInfo>
        <genre authority="marcgt">periodical</genre>
        <genre authority="bibutilsgt">academic journal</genre>
    </relatedItem>
    <abstract>We present a qualitative analysis of the (potentially erroneous) outputs of contextualized embedding-based methods for detecting diachronic semantic change. First, we introduce an ensemble method outperforming previously described contextualized approaches. This method is used as a basis for an in-depth analysis of the degrees of semantic change predicted for English words across 5 decades. Our findings show that contextualized methods can often predict high change scores for words which are not undergoing any real diachronic semantic shift in the lexicographic sense of the term (or at least the status of these shifts is questionable). Such challenging cases are discussed in detail with examples, and their linguistic categorization is proposed. Our conclusion is that pre-trained contextualized language models are prone to confound changes in lexicographic senses and changes in contextual variance, which naturally stem from their distributional nature, but is different from the types of issues observed in methods based on static embeddings. Additionally, they often merge together syntactic and semantic aspects of lexical entities. We propose a range of possible future solutions to these issues.</abstract>
    <identifier type="citekey">kutuzov-etal-2022-contextualized</identifier>
    <identifier type="doi">10.3384/nejlt.2000-1533.2022.3478</identifier>
    <location>
        <url>https://aclanthology.org/2022.nejlt-1.9/</url>
    </location>
    <part>
        <date>2022</date>
        <detail type="volume"><number>8</number></detail>
    </part>
</mods>
</modsCollection>

%0 Journal Article
%T Contextualized embeddings for semantic change detection: Lessons learned
%A Kutuzov, Andrey
%A Velldal, Erik
%A Øvrelid, Lilja
%J Northern European Journal of Language Technology
%D 2022
%V 8
%I Linköping University Electronic Press
%C Linköping, Sweden
%F kutuzov-etal-2022-contextualized
%X We present a qualitative analysis of the (potentially erroneous) outputs of contextualized embedding-based methods for detecting diachronic semantic change. First, we introduce an ensemble method outperforming previously described contextualized approaches. This method is used as a basis for an in-depth analysis of the degrees of semantic change predicted for English words across 5 decades. Our findings show that contextualized methods can often predict high change scores for words which are not undergoing any real diachronic semantic shift in the lexicographic sense of the term (or at least the status of these shifts is questionable). Such challenging cases are discussed in detail with examples, and their linguistic categorization is proposed. Our conclusion is that pre-trained contextualized language models are prone to confound changes in lexicographic senses and changes in contextual variance, which naturally stem from their distributional nature, but is different from the types of issues observed in methods based on static embeddings. Additionally, they often merge together syntactic and semantic aspects of lexical entities. We propose a range of possible future solutions to these issues.
%R 10.3384/nejlt.2000-1533.2022.3478
%U https://aclanthology.org/2022.nejlt-1.9/
%U https://doi.org/10.3384/nejlt.2000-1533.2022.3478

Markdown (Informal)

[Contextualized embeddings for semantic change detection: Lessons learned](https://aclanthology.org/2022.nejlt-1.9/) (Kutuzov et al., NEJLT 2022)

• Contextualized embeddings for semantic change detection: Lessons learned (Kutuzov et al., NEJLT 2022)

ACL

• Andrey Kutuzov, Erik Velldal, and Lilja Øvrelid. 2022. Contextualized embeddings for semantic change detection: Lessons learned. Northern European Journal of Language Technology, 8.