Vaccine-related information is awash on social media platforms like Twitter and Facebook. One party supports vaccination, while the other opposes vaccination and promotes misconceptions and misleading information about the risks of vaccination. The analysis of social media posts can give significant information into public opinion on vaccines, which can help government authorities in decision-making.This paper describes the dataset used in the shared task, and compares the performance of diferent classification that are provax, antivax and last neutral for identifying efective tweets related to Covid vaccines.We experimented with a classification-based approach. Our experiment shows that SVM classification performs well in order to efiective post.We're going to do this because vaccination is an important step for Covid19 so people can easily fix the news about the vaccine and grab their own slot and symptom detection is also playing a important part to arrest the spread of disease.
• Task 1: COVID-19 vaccine stance classification from tweets
The Covid-19 vaccination is of utmost importance to prevent the disease.Here the task is to build a efective classifier to understand the user’s stance in social media.The three classes are as follows: 1. AntiVax - the tweet indicates hesitancy (of the user who posted the tweet) towards the use of vaccines. 2. ProVax - the tweet supports / promotes the use of vaccines. 3. Neutral - the tweet does not have any discernible sentiment expressed towards vaccines or is not related to vaccines • Task 2: Detection of COVID-19 symptom-reporting in tweets
It is important for authorities to identify people who are experiencing COVID-19 symptoms as quickly as possible in order to prevent the spread of the disease. The purpose of this task is to detect tweets that relate to COVID-19 symptoms (e.g., ’fever’, ’cough’). Such tweets are referred to as symptom-reporting tweets.Build an efective classifier for 4-class classification on tweets that can detect tweets that report someone experiencing COVID-19 symptoms. The 4 classes are described below: 1. Primary Reporting - The user (who posted the tweet) is reporting symptoms of himself/herself. 2. Secondary Reporting - The user is reporting symptoms of some friend / relative / neighbour / someone they met. 3. Third-party Reporting - The user is reporting symptoms of some celebrity / third-party person. 4. Non-Reporting - The user is not reporting anyone experiencing COVID-19 symptoms, but talking about symptom-words in some other context. This class includes tweets that only give general information about COVID-19 symptoms, without specifically reporting about a person experiencing such symptoms.
.
• For the Task1 we got 4392 tweets that were labelled with 3 classes.
• For the Task2, there was 1574 tweets that were labelled with 4 classes.
We discuss the overall design and implementation of our approach in this section. We borrow the advantages and capabilities of machine learning algorithms to implement the two abovementioned tasks which are properly discussed in the consequent subsections. 4.1. Task1
COVID-19 vaccine stance classification from tweets : The purpose of this task is to identify the public sentiments toward vaccines based on Twitter data. Model selection, vectorization, and preprocessing are the three steps of the overall process. This phase involves cleaning up of the provided tweets labeled as AntiVax, Provax or Neutral in the training dataset.All the words starting with hashtags or containing multiple spaces, special characters,urls,punctuations or stopwords are firstly trimmed from every tweet. After streaming and lemmatization are applied,the pre processed level tweets are then transform to numeric feature vector using term frequency inverse document frequency. After transforming the unstructured tweet data into numeric structured data.
This numeric structured data was fed into two classifiers: Naive Bayes and SVM(Support Vector Machine). In comparison with these two classifiers, SVM performed better. In this training set, 1676 provax tweets and 1081 antivax tweets are used as training data, and 1635 neutral tweets are used for training and get traing accuracy 0.65.
0 1 2 accuracy macro avg weighted avg precision 0.67 0.62 0.66 0.65 0.65 4.2. Task2
Detection of COVID-19 symptom-reporting in tweets IThe purpose of this task is to identify the symptoms toward Covid-19 based on Twitter data. Model selection, vectorization, and preprocessing are the three steps of the overall process. This phase involves cleaning up of the provided tweets labeled as AntiVax, Provax or Neutral in the training dataset.All the words starting with hashtags or containing multiple spaces, special characters,urls,punctuations or stopwords are firstly trimmed from every tweet.
After streaming and lemmatization are applied,the pre processed level tweets are then transform to numeric feature vector using term frequency inverse document frequency. After transforming the unstructured tweet data into numeric structured data.
This numeric structured data was fed into two classifiers: Naive Bayes and SVM(Support Vector Machine). In comparison with these two classifiers, SVM performed better.In this training set containing 437 primary reporting tweets and 127 Secondary Reporting tweets and 196 Third-Party reporting tweets and 814 tweets are Non-Reporting and get training accuracy 0.69. 0 1 2 3 accuracy macro avg weighted avg
Task 1 - IRMiDis Track results are evaluated using overall accuracy and the macro-F1 score on the three classes as metrics. The result of our submitted automated run for Task 1 is shown in Table 1.
Team ID Subinay-IISERK
Vax_labels.csv
Task 2 - IRMiDis Track results are evaluated using overall accuracy and the macro-F1 score on the four classes as metrics. The result of our submitted automated run for Task 2 is shown in Table 2.
Team ID Subinay-IISERK
Symptoms_labels.csv
Natural Language Processing was used for this work submitted to the IRMiDis Track processing of the tweets. Machine Learning models were used to identify tweets in the training data . We used the NeatText package of NLP for cleaning our tweets. Then techniques like Tokenization and Lemmatization were used to pre-process the tweets. Later, we used a few Machine Learning models to train our dataset among which the SVM(Support Vector Machine) model gave the highest F-score.