EUVIMEDObjevte EUVIMED
Open access · CC BY-NC-ND 4.0

CIÊNCIA DE DADOS APLICADA AO PROCESSAMENTO DE LINGUAGEM NATURAL: UM ESTUDO DA APLICAÇÃO DE ÁLGEBRA DE CLIFFORD EM ANÁLISE DE SENTIMENTOS

André Costa, Alexandre Oliveira

Anais do Congresso Brasileiro de Iniciação Científica · 2026-11-22 · 3 (1) · 3862-3867

DOI: 10.66221/mf08xt58 ↗

At a glance

Journal article in Anais do Congresso Brasileiro de Iniciação Científica, published online on 2026-11-22. Written by André Costa and 1 co-authors. The full text is freely available (open access, CC BY-NC-ND 4.0).

AbstraktBibliografické údajeCiteRelated studies

Abstract (original language)

Natural Language Processing (NLP) faces recurrent challenges in Sentiment Analysis tasks, primarily due to semantic ambiguities, contextual subtleties, and the occurrence of irony. Traditional computational methods frequently employ linear representations or procedural approaches that simplify the geometric structure underlying natural language. This work proposes an alternative approach grounded in Clifford Algebra, modeling text sentences as multivectors in a space with well-defined geometric signatures Cl(p, q, r)$. The methodology consists of extracting adjectives from X (formerly Twitter) posts through linguistic processing (NLTK), followed by lexical polarization (VADER) and TF-IDF vectorization to generate the coefficients and dimensions of the multivector. Four transformation metrics from the multivector to the field of Real numbers were evaluated: the square norm, the general linear functional, the matrix trace via Pauli representations, and the bilinear form. Experimental results with a dataset of 700 posts indicate that all four approaches showed consistency in sentiment classification, achieving an accuracy of 70%, which demonstrates the potential of Clifford Algebra for NLP tasks. Finally, the limitations of the lexical polarizer are discussed, and future perspectives using Portuguese-language datasets are proposed.

Cite

APA

Costa, A., & Oliveira, A. (2026). CIÊNCIA DE DADOS APLICADA AO PROCESSAMENTO DE LINGUAGEM NATURAL: UM ESTUDO DA APLICAÇÃO DE ÁLGEBRA DE CLIFFORD EM ANÁLISE DE SENTIMENTOS. Anais do Congresso Brasileiro de Iniciação Científica, 3(1), 3862-3867. https://doi.org/10.66221/mf08xt58

Vancouver

Costa A, Oliveira A. CIÊNCIA DE DADOS APLICADA AO PROCESSAMENTO DE LINGUAGEM NATURAL: UM ESTUDO DA APLICAÇÃO DE ÁLGEBRA DE CLIFFORD EM ANÁLISE DE SENTIMENTOS. Anais do Congresso Brasileiro de Iniciação Científica. 2026;3(1):3862-3867. doi:10.66221/mf08xt58

BibTeX
@article{Costa2026,
  author = {Costa, André and Oliveira, Alexandre},
  title = {CIÊNCIA DE DADOS APLICADA AO PROCESSAMENTO DE LINGUAGEM NATURAL: UM ESTUDO DA APLICAÇÃO DE ÁLGEBRA DE CLIFFORD EM ANÁLISE DE SENTIMENTOS},
  journal = {Anais do Congresso Brasileiro de Iniciação Científica},
  year = {2026},
  volume = {3},
  number = {1},
  pages = {3862--3867},
  doi = {10.66221/mf08xt58}
}
Start for free →