click here to download project base paper
click here to download project abstract
– ABSTRACT
Sarcasm is an important part of communication. Sarcasm is expressed in words, facial
expressions and can be noticed in the intonation of voice. So in this digital age , sarcastic
comment are passed everyday via tweets, comment sections of different social media outlets
, and even news headlines. Newspapers often seem to use sarcasm in their headlines to grab
the reader’s attention. Some of these headlines can be misunderstood and taken to mean
something different than the original intentions. However, more often than , not the readers
find it difficult to detect the irony within the headlines thus, getting the incorrect idea about
the actual news and further passing on their understanding to their friends and colleagues
.This leads to a need to detect sarcasm especially in the news and on social media. The
sarcasm detection of text has its challenges because text lacks intonations and deflection of
voices that occur when a sarcastic statement is formed vocally by a person’s.
This project focuses on the effect of the different encoding methods used in the text to
extract the feature of machine learning models. A deep learning model is used and the results
are compared. Prior to the feature extraction, pre-processing data techniques such as
tokenization, training set and testing set, and embedding concepts and punctuation are used
by researchers in any work that involves textual analysis. These pre-processing methods are
widely used and accepted and used in this project. Different methods of extracting features
such as Count Vectorizer and word embedding were used in this project. Random Forest ,
and Logistic Regression were the machine learning algorithms used in this project.
Keywords: Sarcasm Detection, Convolutional Neural Networks (CNN), Random forest,
Logistic regression, deep learning.
