Sentimental analysis of COVID-19 vaccine related tweets

Mohannad Rayani, Suboh Alkhushayni, College of Science, Engineering, and Technology, Minnesota State University, Mankato, Mankato, MN 56001

The year 2020 will be remembered in history for the widespread destruction caused by the COVID-19 pandemic. Pharmaceutical companies around the world have started working on the vaccine with many of them being in their final trial stage. However, vaccines take many years before they are available to the general public. In the current scenario, it is expected that it will be available by the end of 2020. Twitter is an online platform where people share their thoughts in the form of tweets. These tweets carry sentiments regarding specific topics organized by hashtags. This study’s main motive is to construct a domain-specific approach that will analyze the sentiments within tweets related to the COVID-19 vaccine. This can be done by gathering COVID-19 vaccine-specific tweets from the Twitter API. The tweets are then processed into a dataset that is suitable for the sentiment classifier algorithms. The dataset is then fed into the sentiment classifier algorithms with customized settings and features. The algorithms are used in both an individual and a hybrid model with n-gram features to compare the accuracy of the two models. Our study shows how a 4-gram feature set affects the accuracy of the hybrid and individual sentiment classifier models.

Additional Abstract Information

Presenter: Mohannad Rayani

Institution: Minnesota State University, Mankato

Type: Oral

Subject: Computer Science

Status: Approved

Time and Location

Session: Oral 8
Date/Time: Tue 5:00pm-6:00pm
Session Number: 814
List other presenters in this same room and session