The rise in online interactions has increased hate speech, making its distinction from offensive language a key challenge. This paper introduces a novel methodology that combines structured-based and learning-based approaches to enhance classification. Traditional and deep learning-based features fail to retain contextual and grammatical meaning. By analyzing 1,700 sentences using constituency parse trees, we identified structural templates for hate speech. An Open Information Extraction System was developed to automate feature extraction using heuristic algorithms. We developed a Linguistic Structure Features Extraction System (LSFES). Our methodology shows a high precision in classification compared to the baseline.

LSFES: A Linguistic Structure Feature Extraction System for Hate Speech and Offensive Language Classification / Alromaema, W. A. M.; Casetti, C. E.; Dragoni, A. F.. - (2025), pp. 311-316. ( 4th IEEE International Conference on Metrology for eXtended Reality, Artificial Intelligence and Neural Engineering, MetroXRAINE 2025 Ancona, IT 22 - 24 October 2025) [10.1109/MetroXRAINE66377.2025.11340418].

LSFES: A Linguistic Structure Feature Extraction System for Hate Speech and Offensive Language Classification

Alromaema W. A. M.
;
Dragoni A. F.
2025-01-01

Abstract

The rise in online interactions has increased hate speech, making its distinction from offensive language a key challenge. This paper introduces a novel methodology that combines structured-based and learning-based approaches to enhance classification. Traditional and deep learning-based features fail to retain contextual and grammatical meaning. By analyzing 1,700 sentences using constituency parse trees, we identified structural templates for hate speech. An Open Information Extraction System was developed to automate feature extraction using heuristic algorithms. We developed a Linguistic Structure Features Extraction System (LSFES). Our methodology shows a high precision in classification compared to the baseline.
2025
9798331502799
File in questo prodotto:
File Dimensione Formato  
Alromaema_LSFES-Linguistic-Structure-Feature_2025.pdf

Solo gestori archivio

Tipologia: Versione editoriale (versione pubblicata con il layout dell'editore)
Licenza d'uso: Tutti i diritti riservati
Dimensione 900.83 kB
Formato Adobe PDF
900.83 kB Adobe PDF   Visualizza/Apri   Richiedi una copia
LSFES_Waleed_Alromaema.pdf

accesso aperto

Tipologia: Documento in post-print (versione successiva alla peer review e accettata per la pubblicazione)
Licenza d'uso: Licenza specifica dell'editore
Dimensione 720.16 kB
Formato Adobe PDF
720.16 kB Adobe PDF Visualizza/Apri

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11566/355413
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 0
  • ???jsp.display-item.citation.isi??? ND
social impact