@inproceedings{ahmadi-zeinali-2025-wordwiz,
title = "{W}ord{W}iz at {S}em{E}val-2025 Task 10: Optimizing Narrative Extraction in Multilingual News via Fine-Tuned Language Models",
author = "Ahmadi, Ruhollah and
Zeinali, Hossein",
editor = "Rosenthal, Sara and
Ros{\'a}, Aiala and
Ghosh, Debanjan and
Zampieri, Marcos",
booktitle = "Proceedings of the 19th International Workshop on Semantic Evaluation (SemEval-2025)",
month = jul,
year = "2025",
address = "Vienna, Austria",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.semeval-1.170/",
pages = "1276--1281",
ISBN = "979-8-89176-273-2",
abstract = "This paper presents our WordWiz system for SemEval-2025 Task 10: Narrative Extraction. We employed a combination of targeted preprocessing techniques and instruction-tuned language models to generate concise, accurate narrative explanations across five languages. Our approach leverages an evidence refinement strategy that removes irrelevant sentences, improving signal-to-noise ratio in training examples. We fine-tuned Microsoft{'}s Phi-3.5 model using both Supervised Fine-Tuning (SFT). During inference, we implemented a multi-temperature sampling strategy that generates multiple candidate explanations and selects the optimal response using narrative relevance scoring. Notably, our smaller Phi-3.5 model consistently outperformed larger alternatives like Llama-3.1-8B across most languages. Our system achieved significant improvements over the baseline across all languages, with F1 scores ranging from 0.7486 (Portuguese) to 0.6839 (Bulgarian), demonstrating the effectiveness of evidence-guided instruction tuning for narrative extraction."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="ahmadi-zeinali-2025-wordwiz">
<titleInfo>
<title>WordWiz at SemEval-2025 Task 10: Optimizing Narrative Extraction in Multilingual News via Fine-Tuned Language Models</title>
</titleInfo>
<name type="personal">
<namePart type="given">Ruhollah</namePart>
<namePart type="family">Ahmadi</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hossein</namePart>
<namePart type="family">Zeinali</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 19th International Workshop on Semantic Evaluation (SemEval-2025)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Sara</namePart>
<namePart type="family">Rosenthal</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Aiala</namePart>
<namePart type="family">Rosá</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Debanjan</namePart>
<namePart type="family">Ghosh</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Marcos</namePart>
<namePart type="family">Zampieri</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Vienna, Austria</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-273-2</identifier>
</relatedItem>
<abstract>This paper presents our WordWiz system for SemEval-2025 Task 10: Narrative Extraction. We employed a combination of targeted preprocessing techniques and instruction-tuned language models to generate concise, accurate narrative explanations across five languages. Our approach leverages an evidence refinement strategy that removes irrelevant sentences, improving signal-to-noise ratio in training examples. We fine-tuned Microsoft’s Phi-3.5 model using both Supervised Fine-Tuning (SFT). During inference, we implemented a multi-temperature sampling strategy that generates multiple candidate explanations and selects the optimal response using narrative relevance scoring. Notably, our smaller Phi-3.5 model consistently outperformed larger alternatives like Llama-3.1-8B across most languages. Our system achieved significant improvements over the baseline across all languages, with F1 scores ranging from 0.7486 (Portuguese) to 0.6839 (Bulgarian), demonstrating the effectiveness of evidence-guided instruction tuning for narrative extraction.</abstract>
<identifier type="citekey">ahmadi-zeinali-2025-wordwiz</identifier>
<location>
<url>https://aclanthology.org/2025.semeval-1.170/</url>
</location>
<part>
<date>2025-07</date>
<extent unit="page">
<start>1276</start>
<end>1281</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T WordWiz at SemEval-2025 Task 10: Optimizing Narrative Extraction in Multilingual News via Fine-Tuned Language Models
%A Ahmadi, Ruhollah
%A Zeinali, Hossein
%Y Rosenthal, Sara
%Y Rosá, Aiala
%Y Ghosh, Debanjan
%Y Zampieri, Marcos
%S Proceedings of the 19th International Workshop on Semantic Evaluation (SemEval-2025)
%D 2025
%8 July
%I Association for Computational Linguistics
%C Vienna, Austria
%@ 979-8-89176-273-2
%F ahmadi-zeinali-2025-wordwiz
%X This paper presents our WordWiz system for SemEval-2025 Task 10: Narrative Extraction. We employed a combination of targeted preprocessing techniques and instruction-tuned language models to generate concise, accurate narrative explanations across five languages. Our approach leverages an evidence refinement strategy that removes irrelevant sentences, improving signal-to-noise ratio in training examples. We fine-tuned Microsoft’s Phi-3.5 model using both Supervised Fine-Tuning (SFT). During inference, we implemented a multi-temperature sampling strategy that generates multiple candidate explanations and selects the optimal response using narrative relevance scoring. Notably, our smaller Phi-3.5 model consistently outperformed larger alternatives like Llama-3.1-8B across most languages. Our system achieved significant improvements over the baseline across all languages, with F1 scores ranging from 0.7486 (Portuguese) to 0.6839 (Bulgarian), demonstrating the effectiveness of evidence-guided instruction tuning for narrative extraction.
%U https://aclanthology.org/2025.semeval-1.170/
%P 1276-1281
Markdown (Informal)
[WordWiz at SemEval-2025 Task 10: Optimizing Narrative Extraction in Multilingual News via Fine-Tuned Language Models](https://aclanthology.org/2025.semeval-1.170/) (Ahmadi & Zeinali, SemEval 2025)
ACL