This section presents a review of the recent literature on the COVID-19 pandemic which emphasizes the importance of effective vaccination for the whole population.

Machine learning and neural networks have applications in difference domains such as aerial image classification [23,24,25,26], face recognition [27], Internet of Things [28,29], healthcare [30,31,32] and sentiment analysis, etc. Manguri et al. [33] stated that the rise of social data on the internet has accelerated. This leads to study in order to obtain access to the data and information for a variety of academic and commercial purposes. The global COVID-19 sickness has now expanded internationally, and social data on the web includes numerous real-life incidents that happened in everyday life. Many people, including media outlets and government institutions, are disseminating the newest information and viewpoints on the coronavirus. The Twitter data was crawled from Twitter social media through a python programming language, and sentiment analysis was performed using the text blob library in python. The evaluation results of sentiment analysis are shown as a graphical representation based on the data. The information originated from Twitter, where it was discovered via the use of a search for two distinct hashtag keywords: (COVID-19 and coronavirus). In another study [34], the authors argued that a global infrastructure to enable both normal and pandemic/epidemic adult vaccination is urgently needed because of the global connections. Since the number of older persons is continually increasing, the need for a framework to propose vaccinations and establish strong platforms to distribute them was obvious. For older individuals, their families, communities, and nations, adult vaccination as a policy has the potential to protect and improve medical, social, and economic results. COVID-19 vaccinations will soon be available, but it is important to remember that currently, a number of vaccines are available that can keep adults healthy.

Meena et al. [35] pointed out that social media talks about healthcare were an excellent starting point for assessing people’s feelings. COVID-19 vaccination was the primary hope of practically every human being on Earth. Many people took to Twitter to express their feelings in response to Russia’s first vaccination announcement. Data from tweets were analyzed for the emotions and psychology of the people and the issue of interest they were discussing. The social emotions were disclosed and displayed using computational approaches and algorithms, such as machine-learned and LDA. Sentiment analysis is a technique for recognizing and categorizing views or feelings represented in the source material. A vast amount of data that is rich in sentiment is generated by various types of social media, such as tweets, status updates, blog posts, and so on. The application of sentiment analysis to this user-generated data may be highly helpful in identifying the perspective of the general population. Because of the existence of slang phrases and misspellings, Twitter sentiment analysis is more complex than conventional sentiment analysis. On Twitter, the maximum number of characters permitted is 140. According to authors, there are two methodologies that are employed for interpreting the sentiment gleaned from the text. These are the knowledge-based approach. Alliheibi [36], mentioned that individuals in Saudi Arabia who had received the COVID-19 vaccination were studied via their tweets. People’s replies were classified using computational lexical-semantic approaches. The findings show that the majority of Saudi Arabians have an unfavorable view of the government’s COVID-19 immunization take-up campaign. According to the findings, the use of data mining applications in government institutions and departments can identify trends that could have an adverse impact on policies and practices, as well as help government institutions make appropriate decisions and adopt reliable and workable policies and procedures.

Yousefinaghani et al. [37] pointed out that COVID-19 vaccinations are the subject of an estimated 4.5 million tweets being analyzed in their investigation. It is possible that Twitter, as it was in this study, may be an effective tool for promoting public health by increasing vaccination uptake and decreasing vaccine resistance. Public health officials might benefit from better knowing vaccine feelings and opinions in order to amplify good postings with supportive language and debunk negative ones with confrontational language that spreads misinformation. Public health organizations may also be able to use Twitter and other media to raise positive messaging and actively minimize negative and opposing messages.

Ezhilan et al. [38] performed a study using a convolutional neural network and a recurrent neural network built for sentiment analysis based on text data related to Twitter data sentiment analysis. CNN and RNN sentiment classifiers performed better than other sentiment classifiers, such as SVM, logistic regression, and Nave Bayes, in terms of accuracy and recall, according to the empirical assessment in this study. Also shown in the study was the performance of general-purpose emotion analyzers such as text blob and Vader. Understanding public opinions regarding coronavirus and COVID-19 helps to detect the rise in dread sentiment and unpleasant feelings, which were important for developing much-needed remedies to stop the rapid spread of the pandemic. The use of exploratory and descriptive text analytics and data visualization methodologies helps to uncover the most basic of ideas. Andrzejczak-Grzadko et al. [39] observed that the Vaccine side effects are widespread, although individuals respond to immunizations in various ways. Manufacturers give a list of their goods’ adverse effects. Adverse responses indicate that immunizations are effective and that the immune system is reacting. It compares the AstraZeneca and Pfizer vaccines’ side effects. These responses were more prevalent after the first dosage of the AstraZeneca vaccination than after the first and second doses of the Pfizer vaccine, although they were less common after the Pfizer formulation. The survey was made available on the internet. It was performed on patients who had been immunized with Pfizer or AstraZeneca vaccines. The participants were questioned about adverse effects such as injection site discomfort, arm pain, muscle pain, headache, fever, chills, and exhaustion after receiving the first and second doses of the vaccinations. A total of 705 persons responded to the survey. Pfizer had vaccinated 196 of them, whereas AstraZeneca had immunized 509. A total of 96.5% of those who received the first dose of the AstraZeneca vaccine had at least one post-vaccination response. All of the adverse effects mentioned in the survey were reported by 17.1% of respondents. Vaccine responses were recorded by 93.9% of those who received the first Pfizer dosage, while just 2% of those who received the second dose suffered all of the adverse events listed in the survey. Most of the subjects had post-vaccinal reactions after the second dose of the Pfizer vaccine: 54.8% had more adverse reactions, and 15.8% had fewer adverse reactions than after the first dose, and 29.4% had the same side effects after the first and second doses of the Pfizer vaccine.

Saeed et al. [40] stated that some people were reluctant to get their children vaccinated because they were afraid of the unknown. The first and second post-vaccination side effects of the Sinopharm COVID-19 vaccine were shown to be common and moderate, predictable, non-serious, and not life-threatening. For the first time, the Sinopharm vaccine’s adverse effects have been evaluated among an age group, and the findings might help lessen public vaccination skepticism. Dubey [41] performed a study to explain. In India, the campaign to prevent COVID-19 began on 16 January 2021. Oxford-Covishield AstraZeneca’s and Bharat Biotech’s Covaxin were two vaccines employed in this campaign. This initiative has already surpassed 600,000 people in its first four days, and the government has declared that it would be increased in the following days to secure residents’ immunity. However, there is still a segment of the population that is skeptical about the COVID-19 vaccine. It was carried out to examine the emotions expressed in India’s tweets about these two vaccinations. While the majority of the public has favorable feelings about these vaccinations, the study indicates that there are also negative feelings about them, which are linked to emotions such as fear and wrath. Dumre et al. [42] performed a statistical and sentiment analysis and observed that people in India have begun developing opinions towards them as a result of the impending availability of a vaccine against COVID-19. An investigation of the attitudes and viewpoints of individuals with respect to vaccinations. Out of 200 participants, 32 doctors and 35 participants were vaccinated. The main objectives were to analyze the response to the survey and draw conclusions with the help of data analysis techniques and performed sentiment analysis on participants’ responses to identify what stops people from getting vaccinated.

Cotfas et al. [43] described that machine learning-based posture detection was used to analyze the one-month time between the initial announcement of a coronavirus vaccine and the first real immunization procedure outside of the limited clinical trials. The best classifier was selected after a thorough evaluation of the performance of a number of different conventional and deep learning methods. The suggested method was able to classify the tweets into three primary categories, namely in favor, against, and neutral, with an accuracy of 78.94%. The authors in [44] analyzed that the tweets were categorized into four different emotions based on their content: fear, sadness, rage, and joy. A pleasant environment was produced in the healthcare authorities by using phrases such as “thank you”, “well”, and “good” instead of terms that instill dread in the minds of those who hear them. In light of these findings, local governments have been pushed to impose fact-checkers on social media to combat misleading propaganda. There has been a lack of research on how to verify and categorize tweets, which has led to a rise in the spread of false information. As a result, the authors used Bert, a unique deep-learning model, to obtain better classification accuracy in comparison to standard models of ML. Bert’s 89% accuracy outperformed other models including LR, SVM, and LSTM, according to the results. The research results helped to clarify public opinion on pandemics and provided a guideline to medical authorities, public, and private sector employees to overcome unnecessary concern during pandemics.