Introduction

According to Iacus, S.M. (2020), COVID-19 is a communicable disease that can be spread through physical touch and or non-physical contact such as through sneezing, coughing, or talking. COVID-19 is the acronym for Corona Virus Disease of 2019. The World Health Organization (WHO) labelled Convid’19 as a global pandemic on March 11th 2020 (Chakraborty, K. et al. 2020). The impact of Convid’19 is social and economical and was reported to be an epidemic on January 30th, 2020 (Guo, Y.R 2019). Convid’19 spread rapidly and led to ill health and even deaths. According to Rustam, F. et al. (2021), the number of death from Convid’19 was over 600 thousand as of May 31st 2020 and this was predicted to increase. Chintalapudi, N. et al. (2021) argued that the world population's mental and physical health has been proven to be related to the Convid’19 pandemic. As of August 27th 2020; more than twenty-four million people had tested positive for Convid’19 worldwide (Chakraborty, K. et al. 2020).

In order to reduce the spread of Convid’19, every country was obligated to develop the best possible way of tracing physical contacts and managing public gathering. One of the most widely adopted way of reducing the spread of the virus was through social distancing and lockdown. Social media has become an important part of human life because It connects people to the outside world. People now heavily depend on posts and tweets published on social media platforms to get news and information. According to Dubey A. D. (2020), social networking traffic saw a dramatic increase during the lockdown.

According to Liu, I.L., Cheung, C.M. and Lee, M.K., (2010), Twitter is the major social media platform where Convid’19 news are timely tweeted and shared. News and thoughts regarding Convid-19 are widely shared on social media (Rustam, F. et al. 2021) and just like a virus, they spread very fast. Koohikamali, M. and Sidorova, A., (2017) argued that due to personal sentiment and bias, most of these news are subjective and the have resulted to misleading information, fake news, and negativity. Just like Convid’19 news gotten from social media are responsible for anxiety due to some misinformation and widespread of fake news and the rapid spread of false information on social has negatively affected mental health (Chintalapudi, N. et al. 2021). To curtail this, it is vital that news spread through mainstream media be checked and regulated. Rustam, F. et al. (2021) thus pointed out that the ability to identify Covid-19 sentiments in tweets would make more educated judgments to be made on how to best handle the spread of misinformation and the Convid’19 pandemic.  Simultaneously, the attention of scholars has been drawn to these issues and one way of understanding and providing solution to the viral misinformation in social media is to use data for Convid’19 sentiment analysis of tweets.

Problem Statement

News that are false and highly biased are rampantly tweeted in Twitter during a global pandemic like the Convid’19 pandemic and they often have extremely negative sentiment. The word "misinfodemic" refers to disinformation that contributes to the transmission of disease (Gyenes, N. and Mina, A.X., 2018). Banerjee, D., (2020) argued that “misinfodemic” was quite common during the Covid’19 outbreak and that scientists largely ignored “misinfodemic” by only focusing on developing a Corona virus vaccine. Extremely biased misinformation is detrimental to the mental health of people (Chintalapudi, N. et al. 2021). “Misinfodemic” have a negative impact on an individual's wellbeing and can lead to societal dysfunction, hysteria (Banerjee, D., 2020) and even suicide.

Aim and Objectives

The aim of this project is to explore, predict and evaluate the sentiments of tweets related to Convid’19. The objectives of this research work are as follows:

1. To better understand and gain insight into the sentiments of Convid’19 tweets.
2. To discover words associated with certain Convid’19 sentiments.
3. To build a machine learning model to predict the sentiment of tweets related to Convid’19.
4. To deploy the machine learning model in a web application to tag the sentiments of tweets related to Convid’19.

Research Questions

The research questions for this project are as follows:

1. What words are mainly associated with a particular sentiment of tweet about Convid’19?
2. Can a machine learning model explain the variation of these words with the sentiments of tweets about Convid’19?
3. Can machine learning be used to flag tweets or posts as being extremely negatively biased so that social media companies can take actions against such tweet/post and maybe the user?

Deliverables

The deliverables of this project are exploratory data analysis, data mining, data munging, modeling, evaluation, model deployment and a project report. The report should thoroughly explain the source of the data used, exploratory data analysis, code implementation, feature selection, model selection as well as model performance.

Relevance

This project is mainly focused on building and deploying a machine learning model that would transform tweets related to Convid’19 and predict the sentiment of the tweets.

Methodology

This project would consider a secondary research, modeling, and deployment. They are further buttressed below:

Secondary research

The secondary research in this project will utilize a systematic approach (Johnson et al., 2016) to review the works of literature. The steps involved in the systematic review of the literature are provided below:

• Step 1: Identify the research questions that can be used for the project.
• Step 2: Identify the keywords that should be used to research the works of literature.
• Step 3: Extract the journals and books that are appropriate for this project.
• Step 4: Write the literature review chapter.

Modeling

This section involves sourcing for and making sense out of data. The steps involved are as follows:

• Data collection
• Data Munging
• Exploratory Data Analysis
• Model selection
• Model optimization
• Model evaluation

Deployment

This section is geared towards packaging the model in a web app and deploying the model. The steps in achieving these are:

• Creating Static web pages
• Styling the web pages
• Creating the backend for the web app to run the model
• Testing the web app
• Deploying the web app to a server
• System testing of the web app

Evaluation

The risk assessment conducted for this project is provided in the table below:

Table 1:  Risk assessment