|
|
Bimonthly Since 1986 |
ISSN 1004-9037
|
|
|
|
|
Publication Details |
Edited by: Editorial Board of Journal of Data Acquisition and Processing
P.O. Box 2704, Beijing 100190, P.R. China
Sponsored by: Institute of Computing Technology, CAS & China Computer Federation
Undertaken by: Institute of Computing Technology, CAS
Published by: SCIENCE PRESS, BEIJING, CHINA
Distributed by:
China: All Local Post Offices
|
|
|
|
|
|
|
|
|
|
Abstract
In today's hyper-digital environment, the number of people using social media platforms has reached an all-time high. The vast majority of people participate in community discussions by posting their ideas and experiences on various social media platforms, such as Twitter, Facebook, and YouTube, amongst others. It is highly vital for both the government and the people in business to conduct research and analysis on the feelings and viewpoints of the general population. This is the motivation behind the active participation of a large number of media organizations during the time of the election in the conduct of a variety of different types of opinion surveys. Using the data from Twitter throughout the time period of the 2019 Lok Sabha election, we have worked in this article to conduct an analysis of the feelings of the people of India during the campaign for that election. Because this is an unsupervised learning problem, we have developed an automatic tweet analyzer that uses the Transfer Learning technique. In our Machine Learning model, we made use of the Linear Support Vector Classifiers approach. Additionally, the Term Frequency Inverse Document Fre-quency (TF-IDF) methodology was utilised in order to manage the textual data that was collected from tweets. In addition to this, we have improved the model's ability to deal with the caustic tweets that are made by some of the users, which is something that the researchers that study this field have not yet taken into consideration.
Keyword
Sentiment Analysis, Sarcasm Detection, Linear SVC, TF-IDF, Political Tweets
PDF Download (click here)
|
|
|
|
|