The global rise of COVID-19 health risk has triggered the related misinformation infodemic. We present the first analysis of COVID-19 misinformation networks and determine few of its implications. Firstly, we analyze the spread trends of COVID-19 misinformation and discover that the COVID-19 misinformation statistics are well fitted by a log-normal distribution. Secondly, we form misinformation networks by taking individual misinformation as a node and similarity between misinformation nodes as links, and we decipher the laws of COVID-19 misinformation network evolution: (1) We discover that misinformation evolves to optimize the network information transfer over time with the sacrifice of robustness. (2) We demonstrate the co-existence of fit get richer and rich get richer phenomena in misinformation networks. (3) We show that a misinformation network evolution with node deletion mechanism captures well the public attention shift on social media. Lastly, we present a network science inspired deep learning framework to accurately predict which Twitter posts are likely to become central nodes (i.e., high centrality) in a misinformation network from only one sentence without the need to know the whole network topology. With the network analysis and the central node prediction, we propose that if we correctly suppress certain central nodes in the misinformation network, the information transfer of network would be severely impacted.

/pdf/deciphering-the-laws-of-social-network-transcendent-covid-19-17z85bicmz.pdf

Deciphering the laws of social network-transcendent COVID-19 misinformation dynamics and implications for combating misinformation phenomena.

This paper presents the findings of an exploratory study on the continuously generating Big Data on Twitter related to the sharing of information, news, views, opinions, ideas, knowledge, feedback, and experiences about the COVID-19 pandemic, with a specific focus on the Omicron variant, which is the globally dominant variant of SARS-CoV-2 at this time. A total of 12,028 tweets about the Omicron variant were studied, and the specific characteristics of the tweets that were analyzed include sentiment, language, source, type, and embedded URLs. The findings of this study are manifold. First, from sentiment analysis, it was observed that 50.5% of tweets had a ‘neutral’ emotion. The other emotions—‘bad’, ‘good’, ‘terrible’, and ‘great’—were found in 15.6%, 14.0%, 12.5%, and 7.5% of the tweets, respectively. Second, the findings of language interpretation showed that 65.9% of the tweets were posted in English. It was followed by Spanish or Castillian, French, Italian, Japanese, and other languages, which were found in 10.5%, 5.1%, 3.3%, 2.5%, and &lt;2% of the tweets, respectively. Third, the findings from source tracking showed that “Twitter for Android” was associated with 35.2% of tweets. It was followed by “Twitter Web App”, “Twitter for iPhone”, “Twitter for iPad”, “TweetDeck”, and all other sources that accounted for 29.2%, 25.8%, 3.8%, 1.6%, and &lt;1% of the tweets, respectively. Fourth, studying the type of tweets revealed that retweets accounted for 60.8% of the tweets, it was followed by original tweets and replies that accounted for 19.8% and 19.4% of the tweets, respectively. Fifth, in terms of embedded URL analysis, the most common domain embedded in the tweets was found to be twitter.com, which was followed by biorxiv.org, nature.com, wapo.st, nzherald.co.nz, recvprofits.com, science.org, and other domains. Finally, to support research and development in this field, we have developed an open-access Twitter dataset that comprises Tweet IDs of more than 500,000 tweets about the Omicron variant, posted on Twitter since the first detected case of this variant on 24 November 2021. 

/pdf/an-exploratory-study-of-tweets-about-the-sars-cov-2-omicron-3h3un476.pdf

An Exploratory Study of Tweets about the SARS-CoV-2 Omicron Variant: Insights from Sentiment Analysis, Language Interpretation, Source Tracking, Type Classification, and Embedded URL Detection

The COVID-19 pandemic hit hard society, strongly affecting the emotions of the people and wellbeing. It is difficult to measure how the pandemic has affected the sentiment of the people, not to mention how people responded to the dramatic events that took place during the pandemic. This study contributes to this discussion by showing that the negative perception of the people of the COVID-19 pandemic is dropping. By negative perception, we mean the number of negative words the users of Twitter, a social media platform, employ in their online posts. Seen as aggregate, Twitter users are using less and less negative words as the pandemic evolves. The conclusion that the negative perception is dropping comes from a careful analysis we made in the contents of the COVID-19 Twitter chatter dataset, a comprehensive database accounting for more than 1 billion posts generated during the pandemic. We explore why the negativity of the people decreases, making connections with psychological traits such as psychophysical numbing, reappraisal, suppression, and resilience. In particular, we show that the negative perception decreased intensively when the vaccination campaign started in the USA, Canada, and the UK and has remained to decrease steadily since then. This finding led us to conclude that vaccination plays a key role in dropping the negativity of the people, thus promoting their psychological wellbeing.

/pdf/negative-perception-of-the-covid-19-pandemic-is-dropping-hx6punjury.pdf

Negative Perception of the COVID-19 Pandemic Is Dropping: Evidence From Twitter Posts.

Garlic-related misinformation is prevalent whenever a virus outbreak occurs. With the outbreak of COVID-19, garlic-related misinformation is spreading through social media, including Twitter. Bidirectional Encoder Representations from Transformers (BERT) can be used to classify misinformation from a vast number of tweets. This study aimed to apply the BERT model for classifying misinformation on garlic and COVID-19 on Twitter, using 5929 original tweets mentioning garlic and COVID-19 (4151 for fine-tuning, 1778 for test). Tweets were manually labeled as ‘misinformation’ and ‘other.’ We fine-tuned five BERT models (BERTBASE, BERTLARGE, BERTweet-base, BERTweet-COVID-19, and BERTweet-large) using a general COVID-19 rumor dataset or a garlic-specific dataset. Accuracy and F1 score were calculated to evaluate the performance of the models. The BERT models fine-tuned with the COVID-19 rumor dataset showed poor performance, with maximum accuracy of 0.647. BERT models fine-tuned with the garlic-specific dataset showed better performance. BERTweet models achieved accuracy of 0.897–0.911, while BERTBASE and BERTLARGE achieved accuracy of 0.887–0.897. BERTweet-large showed the best performance with maximum accuracy of 0.911 and an F1 score of 0.894. Thus, BERT models showed good performance in classifying misinformation. The results of our study will help detect misinformation related to garlic and COVID-19 on Twitter.

/pdf/fine-tuning-bert-models-to-classify-misinformation-on-garlic-qkgry0a5.pdf

Fine-Tuning BERT Models to Classify Misinformation on Garlic and COVID-19 on Twitter

This article deals with the problem of the rapidly increasing COVID-19 infodemic in the world. Thus, there is a need for an effective framework of detecting fake information or misleading news related to COVID-19 virus/disease. To resolve this, we have used a dataset obtained from ConstraintAI'21. The dataset consists of 10,700 tweets and online posts of fake and real news concerning COVID-19. Machine Learning (ML) algorithms compared in this paper to classify the given news or tweet into real or fake are Logistic Regression (LR), K-Nearest Neighbor (KNN), Linear Support Vector Machine (LSVM), Random Forest Classifier (RFC), Decision Tree (DT), Naive Bayes (NB) and Stochastic Gradient Descent (SGD) algorithm. Two feature extraction techniques were used count vectorization and TF-IDF. Deep Learning (DL) algorithms implemented using Adam optimizer are Recurrent Neural Network (RNN), Long Short-Term Memory (LSTM), and Gated Recurrent Unit (GRU). The best testing accuracy was achieved with the LSVM model using TF-IDF feature extraction method followed by Stochastic Gradient Descent classifier with TF-IDF feature extraction technique. LR, DT, and RFC performed better with the Count vectorization feature extraction technique, whereas LSVM, KNN, NB and SGD had better accuracy with TF-IDF feature extraction technique. The LSTM model performed slightly better among the DL algorithms.

COVID-19 Fake News Detection System

/pdf/a-covid-19-rumor-dataset-kx5wkhvak0.pdf

Wenshuo Wang

Papers

A COVID-19 Rumor Dataset