Aspect-Based Sentiment Analysis for Slovene Texts: Models, Lexicons, and Embeddings

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216208%3A11320%2F25%3A9ELXC6VL" target="_blank" >RIV/00216208:11320/25:9ELXC6VL - isvavai.cz</a>

Výsledek na webu

<a href="https://www.scopus.com/inward/record.uri?eid=2-s2.0-85192262138&doi=10.1109%2fIATMSI60426.2024.10503382&partnerID=40&md5=7b9e06c82075cf4b53ca0fe16b9d5aa7" target="_blank" >https://www.scopus.com/inward/record.uri?eid=2-s2.0-85192262138&doi=10.1109%2fIATMSI60426.2024.10503382&partnerID=40&md5=7b9e06c82075cf4b53ca0fe16b9d5aa7</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1109/IATMSI60426.2024.10503382" target="_blank" >10.1109/IATMSI60426.2024.10503382</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Aspect-Based Sentiment Analysis for Slovene Texts: Models, Lexicons, and Embeddings

Popis výsledku v původním jazyce

When performing sentiment analysis on texts mentioning multiple entities, the sentiment towards each of them is not necessarily the same, and it is essential to determine what sentiment applies to each entity separately. In this paper, we study aspect-based sentiment analysis approaches applied to texts in Slovene. We implement three models. In the first model, we use a sentiment lexicon to determine the sentiment of words close to an entity in the same sentence and document and use those as features for a random forest classifier. In the second model, we add a neural model for dependency parsing to the pipeline and construct features based on words close in a sentence's dependency tree instead of sequentially. The third model uses BERT embeddings with a neural classifier to construct embeddings. We evaluate the approaches on the SentiCoref 1.0 corpus of Slovene texts for aspect-based sentiment analysis. © 2024 IEEE.

Název v anglickém jazyce

Aspect-Based Sentiment Analysis for Slovene Texts: Models, Lexicons, and Embeddings

Popis výsledku anglicky

When performing sentiment analysis on texts mentioning multiple entities, the sentiment towards each of them is not necessarily the same, and it is essential to determine what sentiment applies to each entity separately. In this paper, we study aspect-based sentiment analysis approaches applied to texts in Slovene. We implement three models. In the first model, we use a sentiment lexicon to determine the sentiment of words close to an entity in the same sentence and document and use those as features for a random forest classifier. In the second model, we add a neural model for dependency parsing to the pipeline and construct features based on words close in a sentence's dependency tree instead of sequentially. The third model uses BERT embeddings with a neural classifier to construct embeddings. We evaluate the approaches on the SentiCoref 1.0 corpus of Slovene texts for aspect-based sentiment analysis. © 2024 IEEE.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

—

Návaznosti

—

Rok uplatnění

2024

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

IEEE Int. Conf. Interdiscip. Approaches Technol. Manag. Soc. Innov., IATMSI

ISBN

979-835036052-3

ISSN

—

e-ISSN

—

Počet stran výsledku

6

Strana od-do

1-6

Název nakladatele

Institute of Electrical and Electronics Engineers Inc.

Místo vydání

—

Místo konání akce

Gwalior

Datum konání akce

1. 1. 2025

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

—