@inproceedings{hosseini-kivanani-etal-2024-mapping,
title = "Mapping Sentiments: A Journey into Low-Resource {L}uxembourgish Analysis",
author = {Hosseini-Kivanani, Nina and
K{\"u}hn, Julien and
Schommer, Christoph},
editor = "Sousa-Silva, Rui and
Cardoso, Henrique Lopes and
Koponen, Maarit and
Lora, Antonio Pareja and
Seresi, M{\'a}rta",
booktitle = "Proceedings of the First LUHME Workshop",
month = oct,
year = "2024",
address = "Santiago de Compostela, Spain",
publisher = "CLUP, Centro de Lingu{\'i}stica da Universidade do Porto FLUP - Faculdade de Letras da Universidade do Porto",
url = "https://aclanthology.org/2024.luhme-1.3/",
pages = "20--27",
abstract = "Sentiment analysis (SA) plays a vital role in interpreting human opinions across different languages, especially in contexts like social media, product reviews, and other user-generated content. This study focuses on Luxembourgish, a low-resource language critical to Luxembourg`s identity, utilizing advanced deep learning models such as BERT, RoBERTa, LuxemBERTand LuxGPT-2. These models were enhanced with transfer learning, active learning strategies, and context-aware embeddings, enabling effective Luxembourgish processing. These models further improved with context-aware embeddings and were able to accurately detect sentiments, categorizing news comments into positive, negative, and neutral sentiments. Our approach highlights the significant role of human-in-the-loop (HITL) methodologies, which refine model accuracy by aligning automated analyses with human judgment. The findings indicate that LuxembBERT, especially when enhanced with the HITL method involving feedback from 500 and 1000 annotated sentences, outperforms other models in both binary (positive vs. negative) and multi-class (positive, neutral, and negative) classification tasks. The HITL approach not only refined model accuracy but also provided substantial improvements in understanding and processing sentiments and sarcasm, often challenging for automated systems. This study establishes the basis for future research to extend these methodologies to other underresourced languages, promising improvements in Natural Language Processing (NLP) applications across diverse linguistic landscapes."
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="hosseini-kivanani-etal-2024-mapping">
<titleInfo>
<title>Mapping Sentiments: A Journey into Low-Resource Luxembourgish Analysis</title>
</titleInfo>
<name type="personal">
<namePart type="given">Nina</namePart>
<namePart type="family">Hosseini-Kivanani</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Julien</namePart>
<namePart type="family">Kühn</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Christoph</namePart>
<namePart type="family">Schommer</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2024-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the First LUHME Workshop</title>
</titleInfo>
<name type="personal">
<namePart type="given">Rui</namePart>
<namePart type="family">Sousa-Silva</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Henrique</namePart>
<namePart type="given">Lopes</namePart>
<namePart type="family">Cardoso</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Maarit</namePart>
<namePart type="family">Koponen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Antonio</namePart>
<namePart type="given">Pareja</namePart>
<namePart type="family">Lora</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Márta</namePart>
<namePart type="family">Seresi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>CLUP, Centro de Linguística da Universidade do Porto FLUP - Faculdade de Letras da Universidade do Porto</publisher>
<place>
<placeTerm type="text">Santiago de Compostela, Spain</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Sentiment analysis (SA) plays a vital role in interpreting human opinions across different languages, especially in contexts like social media, product reviews, and other user-generated content. This study focuses on Luxembourgish, a low-resource language critical to Luxembourg‘s identity, utilizing advanced deep learning models such as BERT, RoBERTa, LuxemBERTand LuxGPT-2. These models were enhanced with transfer learning, active learning strategies, and context-aware embeddings, enabling effective Luxembourgish processing. These models further improved with context-aware embeddings and were able to accurately detect sentiments, categorizing news comments into positive, negative, and neutral sentiments. Our approach highlights the significant role of human-in-the-loop (HITL) methodologies, which refine model accuracy by aligning automated analyses with human judgment. The findings indicate that LuxembBERT, especially when enhanced with the HITL method involving feedback from 500 and 1000 annotated sentences, outperforms other models in both binary (positive vs. negative) and multi-class (positive, neutral, and negative) classification tasks. The HITL approach not only refined model accuracy but also provided substantial improvements in understanding and processing sentiments and sarcasm, often challenging for automated systems. This study establishes the basis for future research to extend these methodologies to other underresourced languages, promising improvements in Natural Language Processing (NLP) applications across diverse linguistic landscapes.</abstract>
<identifier type="citekey">hosseini-kivanani-etal-2024-mapping</identifier>
<location>
<url>https://aclanthology.org/2024.luhme-1.3/</url>
</location>
<part>
<date>2024-10</date>
<extent unit="page">
<start>20</start>
<end>27</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Mapping Sentiments: A Journey into Low-Resource Luxembourgish Analysis
%A Hosseini-Kivanani, Nina
%A Kühn, Julien
%A Schommer, Christoph
%Y Sousa-Silva, Rui
%Y Cardoso, Henrique Lopes
%Y Koponen, Maarit
%Y Lora, Antonio Pareja
%Y Seresi, Márta
%S Proceedings of the First LUHME Workshop
%D 2024
%8 October
%I CLUP, Centro de Linguística da Universidade do Porto FLUP - Faculdade de Letras da Universidade do Porto
%C Santiago de Compostela, Spain
%F hosseini-kivanani-etal-2024-mapping
%X Sentiment analysis (SA) plays a vital role in interpreting human opinions across different languages, especially in contexts like social media, product reviews, and other user-generated content. This study focuses on Luxembourgish, a low-resource language critical to Luxembourg‘s identity, utilizing advanced deep learning models such as BERT, RoBERTa, LuxemBERTand LuxGPT-2. These models were enhanced with transfer learning, active learning strategies, and context-aware embeddings, enabling effective Luxembourgish processing. These models further improved with context-aware embeddings and were able to accurately detect sentiments, categorizing news comments into positive, negative, and neutral sentiments. Our approach highlights the significant role of human-in-the-loop (HITL) methodologies, which refine model accuracy by aligning automated analyses with human judgment. The findings indicate that LuxembBERT, especially when enhanced with the HITL method involving feedback from 500 and 1000 annotated sentences, outperforms other models in both binary (positive vs. negative) and multi-class (positive, neutral, and negative) classification tasks. The HITL approach not only refined model accuracy but also provided substantial improvements in understanding and processing sentiments and sarcasm, often challenging for automated systems. This study establishes the basis for future research to extend these methodologies to other underresourced languages, promising improvements in Natural Language Processing (NLP) applications across diverse linguistic landscapes.
%U https://aclanthology.org/2024.luhme-1.3/
%P 20-27
Markdown (Informal)
[Mapping Sentiments: A Journey into Low-Resource Luxembourgish Analysis](https://aclanthology.org/2024.luhme-1.3/) (Hosseini-Kivanani et al., LUHME 2024)
ACL