Integrating Technical Analysis into Sentiment Analysis: An ASTE Framework for Electric Car Purchase Decision Support Based on LLMs and Semantic BNF

FilesOpen
Identifiers
URI: http://hdl.handle.net/10952/9362



Statistics
View Usage Statistics
Metadata
Show full item record
Author/s
Carrasco Aguilar, Álvaro; Carmona Martínez, María de las Mercedes; Parra Meroño, María Concepción; Souto Romero, Mar
Date
2025-03-04
Discipline/s
Administración y Dirección de Empresas
Marketing
Ingeniería, Industria y Construcción
Subject/s
Universidad
Univeristy
Repositorio
Repository
Marketing
Aspect sentiment triplet extraction
Large language models
Semantic Backus–Naur Form
Fuzzy linguistic models
Electric vehicles
Expert reviews
Abstract
This study enhances Aspect Sentiment Triplet Extraction (ASTE) by integrating Large Language Models (LLMs) and a semantic Backus-Naur Form (BNF) framework to analyze expert reviews on electric vehicles. A 2-tuple fuzzy linguistic model improves sentiment representation, ensuring interpretability. Applied to YouTube reviews of top-selling EVs, the method effectively structures technical and qualitative data for decision support.
Collections
Attribution-NonCommercial-NoDerivatives 4.0 Internacional
Except where otherwise noted, this item's license is described as Attribution-NonCommercial-NoDerivatives 4.0 Internacional