NORMA eResearch @NCI Library

Transformer based Detection of Sarcasm and it’s Sentiment in Textual Data

Gosavi, Shubham Ram (2022) Transformer based Detection of Sarcasm and it’s Sentiment in Textual Data. Masters thesis, Dublin, National College of Ireland.

[thumbnail of Master of Science]
PDF (Master of Science)
Download (1MB) | Preview
[thumbnail of Configuration manual]
PDF (Configuration manual)
Download (2MB) | Preview


Today data has become the most valuable aspect as it is driving all the major decision making process all around the globe. The biggest platform of data generation is social media where people openly convey their likes and dislikes, and using this data many organizations make business driving decisions. The field of Sentiment Analysis is one such application which helps classify a review of any service or product and help companies to improve their line of services. But automatically classifying sarcasm, which is widely used on the internet, is a challenge. This research proposes a novel approach, Bidirectional Encoder Representations from Transformers (BERT) which is a transformer-based neural network approach to detect sarcasm in the input text and sentiment behind the text with the help of Valence Aware Dictionary for Sentiment Reasoning (VADER). To achieve this, News Headline data originally sourced from website TheOnion will be incorporated with the transformer architecture which predicts if a particular text is sarcastic or not and also the sentiment behind the text using VADER. The analysis of this model shows that BERT comes out to be the best performing model with 77% accuracy in detecting sarcasm and VADER is able to classify each textual sentiment into positive and negative while evaluating that being sarcastic the sentiment seen is mostly negative. The proposed research can be incorporated to build classification systems by organizations who mainly deal in e-commerce to classify the reviews they get on the products and also understand the sentiment behind every review of their customer to improve product or business quality.

Item Type: Thesis (Masters)
Subjects: Q Science > QA Mathematics > Electronic computers. Computer science
T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science
P Language and Literature > P Philology. Linguistics > Computational linguistics. Natural language processing
Z Bibliography. Library Science. Information Resources > ZA Information resources > ZA4150 Computer Network Resources > The Internet > World Wide Web > Websites > Online social networks
T Technology > TK Electrical engineering. Electronics. Nuclear engineering > Telecommunications > The Internet > World Wide Web > Websites > Online social networks
Divisions: School of Computing > Master of Science in Data Analytics
Depositing User: Tamara Malone
Date Deposited: 26 Jan 2023 14:34
Last Modified: 03 Mar 2023 11:59

Actions (login required)

View Item View Item