EXPLORING A NOVEL HEXACO PERSONALITY TRAITS ON TWITTER: AN ENSEMBLE-BASED NLP METHODOLOGY

Natural Language Processing (NLP) plays a crucial role in analyzing Twitter data to introduce an automated HEXACO model. Analyzing personality traits from social media data, particularly on platforms like Twitter, presents unique challenges due to the brevity, informal language, and rapid evolutio...

Full description

Saved in:
Bibliographic Details
Main Authors: Tanvi Desai, Divyakant Meva
Format: Article
Language:English
Published: Institute of Mechanics of Continua and Mathematical Sciences 2025-01-01
Series:Journal of Mechanics of Continua and Mathematical Sciences
Subjects:
Online Access:https://jmcms.s3.amazonaws.com/wp-content/uploads/2025/01/18081432/jmcms-2501020-Exploring-a-Novel-Hexaco-Personality-Traits-TD-DM.pdf
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Natural Language Processing (NLP) plays a crucial role in analyzing Twitter data to introduce an automated HEXACO model. Analyzing personality traits from social media data, particularly on platforms like Twitter, presents unique challenges due to the brevity, informal language, and rapid evolution of linguistic expressions. To overcome these drawbacks, this research presents a methodological framework for investigating a novel HEXACO personality trait using Twitter tweets. The HEXACO model encompasses Honesty-Humility, Emotionality, Extraversion, Agreeableness, Conscientiousness, and Openness to Experience, offering a comprehensive basis for personality analysis. Our approach integrates advanced NLP techniques across key phases: preprocessing, feature extraction, feature selection, and final detection. Preprocessing involves tokenization, stop word removal, and stemming to standardize data quality. Feature extraction leverages contextual Term Frequency-Inverse Document Frequency (TF-IDF), and Global Vectors for Word Representation (GloVe) embeddings models to capture semantic and contextual information from tweets. Feature selection employs the Hybrid Kepler Inspired Secretary Bird (HKISP) algorithm, a combination of the Kepler Optimization Algorithm (KOA) and Secretary Bird Optimization (SBO). The final detection phase utilizes a weighted ensemble voting model comprising Artificial Neural Networks (ANN), Random Forest (RF), and kNearest Neighbours (k-NN) classifiers to enhance predictive accuracy and model robustness. The proposed technique achieved a classification Accuracy of 98.067% and a Hamming loss of 1.933%, which is proved to be superior to the existing models based on the obtained experimental findings.
ISSN:0973-8975
2454-7190