”Sarcasm Detection using Dilbert and Albert: An In-Depth Comparative Analysis with Bert”

Wadhwani, Rohit Gopal (2023) ”Sarcasm Detection using Dilbert and Albert: An In-Depth Comparative Analysis with Bert”. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (4MB) | Preview
Preview	PDF (Configuration manual) Download (5MB) | Preview

Abstract

In today’s communication landscape, sarcasm has become a prevalent way of interacting with one another. However, its written form presents a challenge as the absence of tone makes understanding sarcastic comments difficult. Accurately detecting sarcasm in written text is crucial for grasping the true sentiment behind the words. This ability will not only help people to find the true intention of the headlines but also will aid companies in improving their services for customers and will also help identify business gaps and driving exponential growth. To address this, machine learning models, especially transformers, have been more efficient in sarcasm detection. For this research, we have compiled a large dataset of sarcastic and non-sarcastic news headlines from 2 sources: the onion and huffingtonpost, as headlines are a prime training ground for sarcasm detection. Utilizing transformer-based models like ALBERT and DILBERT, we aim to compare their performance with BERT and give a detailed analysis for the same. While retaining high accuracy of 97%, and 98% and in sarcasm detection, comparing BERT, ALBERT, and DILBERT enables model selection that is optimized based on variables including computing resources, scalability, and training efficiency. I hope to offer fresh approaches to improve sarcasm detection precision and advance NLP techniques for textual content analysis by examining the strengths and shortcomings of each model. Finally, the research findings will contribute to more accurate sarcasm recognition and improved text understanding in a variety of applications.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Muslim Jameel, Syed UNSPECIFIED
Subjects:	Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science H Social Sciences > HM Sociology > Information Science > Communication P Language and Literature > P Philology. Linguistics > Computational linguistics. Natural language processing
Divisions:	School of Computing > Master of Science in Data Analytics
Depositing User:	Tamara Malone
Date Deposited:	08 Jan 2025 18:54
Last Modified:	08 Jan 2025 18:54
URI:	https://norma.ncirl.ie/id/eprint/7289

Actions (login required)

View Item