top of page

Content

Our Research on Vaccine Misinformation is featured on MSU 

"Researchers of the college are dedicated to studying how people interact with their technology and the ways in which they can do so safely and effectively. Their goal is to bring more awareness to students and faculty about the importance of cybersecurity in order to ensure confidentiality, integrity and accessibility when it comes to using various forms of technology."

Our Research on Vaccine Misinformation was Mentioned on NPR

This recent article that was published on NPR explains our use of Artificial intelligence to identify misinformation surrounding the Covid-19 pandemic and vaccines. 

New Publication: The 2022 IEEE International Conference on Digital Health

Dr. Argyris, Nan Zhang, Bidhan Bashyal, and Pang-Ning Tan will be presenting their research paper, "Using Deep Learning to Identify Linguistic Features that Facilitate or Inhibit the Propagation of Anti- and Pro-Vaccine Content on Social Media" with an acceptance rate of 33% at the IEEE International Conference on Digital Health. The event begins July 11th and will be hosted at UPC Terrassa Campus in Barcelona, Spain.

Screen Shot 2022-05-02 at 5.57.10 PM.png

Our Positive Social Influence Lab Presented at the MSU Engineering Graduate Symposium

Our computer science student Bidhan Bashyal presented at the MSU 2022 engineering graduate research conference on how to utilize deep learning to discover linguistic factors that favor or prevent the propagation of anti- and pro-vaccine messages on social media.

Our Lab Presented at the MSU Science Festival - Introduce the Labeling Process to The Public 

The Positive Social Influence Lab was represented by primary speaker Mimi Lui this April when she shared the lab’s process of coding tweets for the social media vaccine hesitancy project at the MSU Science Festival. The lab has been working to create a code book for labelling the thousands of tweets which are being captured through Twitter for use in building an algorithm that can identify tweets that are pro-,anti-, or neutral about vaccines. It was a great opportunity for the lab to share with the public what we are doing to fight online misinformation about vaccines.

Trifecta Talk: learn how a multidisciplinary team work together and make impacts in public health

#3

Screen Shot 2021-07-30 at 2.02.55 PM.png
In-depth: How Social Media Influences Opinions on COVID-19 Vaccines

Source: KSHB.com

#7

In-depth: How Social Media Influences Opinions on COVID-19 Vaccines

Does the information people view on social media truly influence their real-world actions? Young Anna Argyris, Ph.D., an assistant professor in the Media and Information Department at Michigan State University, is working to answer that question in a KSHB-TV interview. Dr. Argyris's advice is to not believe everything posted on social media and find the true source of a post before taking it to heart.

Ask the Expert: Social Media’s Impact on Vaccine Hesitancy

Source: MSU TODAY

#8

Ask the Expert: Social Media’s Impact on Vaccine Hesitancy

Though COVID-19 vaccines are widely available, many people in the United States have not received the vaccine and don’t plan on getting one. Assistant professor Young Anna Argyris, in the College of Communication Arts and Sciences, sheds light on the data and social media influences behind vaccine hesitancy.

Spotlight Winner: Michigan State University (United States)

Source: Google.com

#9

(Google OMC) Spotlight Winner: Michigan State University (United States)

The Online Marketing Challenge (OMC) of Google is a unique opportunity for students to get real-world experience creating and executing online marketing campaigns. Of course, the online marketing challenge is for a non-profit and partnering student who came up as a team member. Participants are from all over the world, and only very few teams will be awarded for their perfect performance. Katie Compo, Zihao Gao, and Dakshaini (Daki) Ravinder, pictured with their professor Young Anna Argyris, from Michigan State University, are the spotlight winner for OMC Spring 2020.

Young.png
Professor Increases HPV Vaccination Rates for Underserved Populations

Source: msu.edu

#10

Professor Increases HPV Vaccination Rates for Underserved Populations

Young Anna Argyris, Ph.D., assistant professor of Media and Information, has been selected as the 2019 recipient of the Diversity Research Network Program Launch Awards Program (DRN-LAP) for researching the influence of these messages on mothers deciding whether or not to vaccinate their children. 

hrcc_hero.jpg
Faculty Feature of the Month by Health & Risk Communication Center

Source: msu.edu, Facebook and Twitter

#11

Faculty Feature of the Month by Health & Risk Communication Center

Dr. Young Anna Argyris, Assistant Professor in the Department of Media and Information, is this month's feature of HRCC. Her research focuses on the organizational use of social media to connect disparate parties, using information technology as a decision-making aid. She also focuses on health misinformation and social media influences.

Orginal Page

#1. Quote From NPR Regarding Our Vaccine Misinformation Research

Contents
#1
#2

"A team at Michigan State University manually sorted over 10,000 tweets to pro-vaccine, anti-vaccine, neutral and irrelevant as training data. The team then used the training data to build a tool that sorted over 120 million tweets into these buckets. For the automatic sorting to remain relatively accurate as the social conversation evolves, humans have to keep annotating new tweets and feed them the training set, Pang-Ning Tan, a co-author of the project, told NPR in an email."

#2. Our Student Present at the MSU Engineering Graduate Symposium

Our computer science student Bidhan Bashyal presented at the MSU 2022 engineering graduate research conference on how to utilize deep learning to discover linguistic factors that favor or prevent the propagation of anti- and pro-vaccine messages on social media.

Screen Shot 2022-05-02 at 6.18.50 PM.png

#3. MSU Science Festival - Introduce the Labeling Process to The Public 

The Positive Social Influence Lab was represented by primary speaker Mimi Lui this April when she shared the lab’s process of coding tweets for the social media vaccine hesitancy project at the MSU Science Festival. The lab has been working to create a code book for labelling the thousands of tweets which are being captured through Twitter for use in building an algorithm that can identify tweets that are pro-,anti-, or neutral about vaccines. It was a great opportunity for the lab to share with the public what we are doing to fight online misinformation about vaccines.

#4. TRIFECTA TALK:Vaccine Informatics Systems and the Impact of Vaccine Debate on Immunization Rates

#3
#4

#6. Argyris and Stern receive NIH Grant to Analyze Impact of Vaccine Misinformation

ComArtSci’s Dr. Young Anna Argyris (PI) and Dr. Michael Stern (Co-I) were recently awarded an R21 grant from the National Institutes of Health (NIH) to study how vaccine misinformation affects COVID-19, flu, and HPV immunization rates. The team will develop a public barometer of the quality and quantity of vaccine debates on social media and explore the impact of that debate on public health.

"This project is innovative because it proposes a robust co-teaching framework for addressing potentially noisy human annotations of vaccine posts. It also proposes a statistical modeling technique that involves heterogenous metrics obtained from a multi-method approach for hypothesis testing,” said Dr. Argyris. “These innovations are timely and urgent as the current time presents a rare opportunity to identify the impact of vaccine debate on public health during the onset of a global pandemic,” continued Dr. Argyris. “The expected outcomes are significant because it will provide a methodological breakthrough in uncovering the reasoning behind vaccine-refusals.”

Joining Dr. Argyris and Dr. Stern on the grant are Pang-Ning Tan (MPI) from the College of Science and Engineering at MSU, Jiying Ling (Co-I) from the College of Nursing at MSU, Robert Jacobsen (consultant) of the Mayo Clinic, and Nada Ganesh (consultant) of NORC at the University of Chicago. The grant ($348,310) supports two years of research (August 2021 to July 2023). 

The project team will collect one million social media posts and develop a deep learning algorithm to identify and classify pro- and anti-vaccine posts. They will then pair the results of social media data analysis with algorithm surveys of U.S. adults to better understand the impact of the vaccine debate on public health outcomes. 

“I am glad to have received this award because the expected outcomes from this project will have a direct positive impact on public health,” said Dr. Argyris. “I have been conducting research on vaccine hesitancy since 2018, so this award will expand and further enable my research activities.”

#7. State Officials Concerned about Vaccine Misinformation on Social Media

LANSING — State officials say they’re concerned about misinformation on coronavirus vaccines on social media as COVID-19 cases continue to rise in Michigan. 

Screen Shot 2021-07-30 at 3.07.34 PM.png

Vaccination misinformation on social media can range from unfounded concerns that the vaccine affects fertility to more interesting takes, such as that 5G cell phone towers can exacerbate COVID-19 symptoms.

Michigan has the nation’s highest COVID-19 rate, according to the state Department of Health and Human Services. 

“We’re very concerned about the recent increase in cases, announcing over 5,000 new cases a day,” said Joneigh Khaldun, the chief medical executive and chief deputy director of the department. 

“Unfortunately, there are a lot of rumors and misinformation going around about the vaccine,” she said. “The vaccine is safe.” 

Attorney General Dana Nessel said people who otherwise would have gotten vaccinated may decide not to due to misinformation spread on social media.

“The request that we would make, and that we have made, to the social media outlets is if you see information that you know is factually inaccurate about the vaccine, to remove it from your platform when asked to do so,” Nessel said. “And certainly not to engage in the process of helping to disseminate it.

“You can’t scream ‘fire’ in a crowded theatre. That’s not your First Amendment right. I don’t believe you should be able to willfully, knowingly disseminate misinformation about the vaccine because you are contributing to people dying,” Nessel said.

Research has shown that exposure to misinformation about vaccines is correlated with lower vaccination rates, said Young Anna Argyris, an assistant professor in the Department of Media and Information at Michigan State University. 

 She said although there is an apparent correlation between anti-vaccination messages and people’s decisions not to get the shots, researchers haven’t proven a direct cause-and-effect relationship.

However there are technologies that can be used to identify vaccine misinformation.

“Technically, it’s a really, really complicated task,” Argyris said. “Think about the number of posts that contain misinformation that are posted every day. Literally millions. We cannot screen them manually.”

She said she’s developed an algorithm — a set of mathematical processes or rules — that is 96% accurate in detecting misinformation.

Even so, the other 4% equals millions of posts that can “cause genuine harm,” Argyris said. Detecting misinformation becomes even more difficult when it comes to images and video.

Those who post misinformation about the vaccine may use different spellings of words, such as “vaccine” or “COVID-19,” to evade detection. 

Khaldun said COVID-19 cases have been rising in Michigan, in part due to “pandemic fatigue,” or people simply being tired of social distancing, as well as misinformation about the virus.

Misinformation about potential side effects could keep people from getting vaccinated, although most side effects are minor, she said. 

“They tend to primarily be mild pain at the injection site, low-grade fever, chills, particularly in younger individuals,” Khaldun said. “And that’s actually to be expected as a result of your body getting primed to be able fight the real virus if they come into contact with it.”

Khaldun said vaccines have proven effective despite misinformation.

“The vaccines are still more than 95% effective in preventing COVID-19,” she said. “And even if you do get COVID-19 after you’ve been fully vaccinated, your likelihood of getting hospitalized or losing your life is extremely low.”


“But I think it’s just really important in that we get appropriate facts out there so that people can learn and understand,” she said. 

#5
#11

#8. In-depth: How Social Media Influences Opinions on COVID-19 Vaccines

OVERLAND PARK, Kan. — When Andrea, a mother of two from Blue Springs, Missouri, scrolls through her social media feeds, there’s always something about COVID-19.

“I just kinda keep scrolling,” she explained.

She doesn’t give much credence to what she reads about COVID-19 vaccinations on social media.

“People are going to say what they want on social media and that’s not always necessarily factual,” added Jordan, an avid social media user.

So, does the information people view on social media truly influence their real-world actions?

Young Argyris, Ph.D., an assistant professor in the Media and Information Department at Michigan State University, is working to answer that question.

She first got involved in the research after giving birth and realizing how much misinformation existed online about childhood immunizations.

Now, her attention is turning toward the COVID-19 vaccine.

A machine-learning algorithm her team developed is analyzing posts and Tweets; categorizing them as either pro-vaccine, neutral or anti-vaccine.

“Very few studies have shown the causal relationship between how we engage with anti-vax messages and our own immunization rate” Argyris explained.

Her study is in the early stages, but she’s been able to draw some takeaways. First, visual posts like memes and pictures draw the most attention.

The interview episode on Dr. Young Anna Argyris about the social media's influence on vaccines.

“Visual stimuli have a greater impact on our information-processing and decision-making,” she explained.

Argyris pointed out many posts spreading misinformation focus on personal and immediate wins. Those views gain traction when friends, family and other trustworthy peers share the posts on social media.

Conversely, posts with scientifically proven vaccine information highlight long-term, societal benefits, which have less personal and immediate appeal.

“This is more a medical doctor wearing a white gown delivering a lecture,” Argyris gave an example.

She said many posts focused on misinformation misspell words or rely on graphics that can sneak past filters dedicated to flagging and removing posts with certain words. She said most cyber filters look for the word “vax,” so posts spreading misinformation now use “vacs” or “vaks” to go undetected.

“It’s very challenging to completely keep our cyberspace free from misinformation,” Argyris admitted.

Her advice is to not believe everything posted on social media and find the true source of a post before taking it to heart.

#6

#9. Ask the Expert: Social Media’s Impact on Vaccine Hesitancy

How do visuals on social media influence people’s likelihood to get vaccinated?

Visuals on social media like memes, videos, photos, posters and emojis are processed faster, accepted without being questioned, and remembered for a longer period than text posts. Especially since the visual often includes a personalized dramatization of vaccine injuries — like an individual having a life-threatening seizure after receiving a vaccine.

Since social media has rapidly grown as a source of news, more and more people are obtaining health information from social media. As a result, visual messaging on social media has significant associations with people’s intentions to get vaccinated — not only against COVID-19, but also for other immunizations. Specifically, our latest study has shown that individuals’ engagement with anti-vaccine messages on social media has a negative impact on their intentions to get vaccinated, while their engagement with pro-vaccine messages has no significant association.

Young.png

Assistant professor Young Anna Argyris, in the College of Communications Arts and Sciences.

What propaganda techniques are used by anti-vaccination groups to influence the conversation around vaccines? Is this leading to more vaccine hesitancy?

Anti-vaccination groups use all the four propaganda techniques known to be effective in political campaigns. They define the pressing issue as vaccine safety/injuries and inefficacy and blame pharmaceutical companies for “cutting corners” to rapidly produce vaccines. They also make moral judgements by suggesting a coalition between corrupted politicians and profit-driven health care industries and recommend rejecting vaccines as a remedy to this problem.

How can we frame the message around vaccines on social media to encourage higher participation in vaccination efforts?

Prior studies on health communication have shown the importance of emphasizing the benefits of taking a health behavior rather than portraying the harms of refusing to take the health behavior. Also, focusing on the immediate and personalized benefits have been found more effective than distant, societal benefits. The revised mask mandate

by Centers for Disease Control and Prevention is in alignment with these findings: Make clear the tangible, immediate and personal benefits of getting COVID-19 vaccines such as, “You can finally take off your masks as long as you get vaccinated!”

Conversely, focusing on societal, long-term benefits, such as the creation of herd immunity, has been known to be ineffective for encouraging individuals to take health behaviors. A well-known example of this communication strategy is to encourage safe sex practices during the HIV epidemic. Instead of focusing on the severe consequences of not using those health practices, communication specialists emphasized the benefits of practicing safe sex — you can enjoy your life freely if you practice safe sex.

In short, the rhetoric for encouraging immunizations on social media should emphasize immediate and personalized benefits of taking the vaccines, rather than long-term protective or societal benefits. The effectiveness will also magnify if these benefits can be visually framed in photos, videos, memes, and/or posters, for the augmented persuasiveness of visual stimuli than textual narratives.

How can social media users tell the difference between high quality vaccination information on social media and propaganda or misinformation?

Discerning accurate information from misinformation is a challenge that individuals may not be able to completely resolve. Social media puts us in a bubble called, “Echo chambers” where we are surrounded by like-minded individuals who reinforce our own existing views rather than being challenged by different views. Studies have shown that debiasing individuals especially from anti-vaccine beliefs is an extremely challenging task because health beliefs are deeply ingrained in our cultural backgrounds, political/religious beliefs and lifestyle choices. Thus, it is recommended to prevent populations that are especially vulnerable and susceptible to health misinformation from being exposed to it in the first place. It is essential to suppress the propagation of vaccine misinformation via social media. These solutions can be embedded in tools like fact-checkers installed in our web browsers that warn readers if the information to be presented is likely to be false.

One of your projects focused on mothers in particular. What are the challenges mothers face in making decisions about whether to get their children vaccinated for COVID-19?

Mothers make over 90% of health decisions for children on their own or jointly with their children and/or partners. In any case, even though the gender roles are changing in this society, mothers still remain the key decision-makers for health-related issues for children. We need to recognize the intensity of stress that mothers feel when they must make decisions that directly affect their children’s well-being. In such emotion-laden circumstances, individuals’ tendency to choose a status quo option — inaction, or refusing to take any action at all — increases disproportionally. So, they end up with the non-decision, “let’s wait and see,” which is manifested in “vaccine hesitancy.”  The intense stress that mothers experience when making vaccination decisions is a crucial factor that increases their tendency to delay or deny immunizations for their children.

You designed a machine learning algorithm to detect anti-vaccination messages online. Can you tell us a little about how that works? How else could this technology be used?

Our deep-learning anti-vaccine detector recognizes multimodal content in social media posts, including text (comments and responses), hashtags and visual elements with 97% of accuracy. To the best of our knowledge, this accuracy has thus far been the highest reported in the literature on anti-vaccine detection.  And I dare to say the 97% of accuracy is higher than human’s detection accuracy. In my experience, humans frequently make errors when they must classify thousands of social media posts due to fatigue they experience after many hours of repeating the same classification task. Our algorithm doesn’t get tired.

In addition, this algorithm can be easily plugged into a web browser to function as a fact-checker that sends a warning message to the user when they are about to be exposed to vaccine misinformation.

What reasons do most people give for not getting vaccinated? Whether based in science or fact, are they sharing these reasons on social media?

Anti-vaccine messages have diverse themes, all of which converge to vaccine safety/injuries and the conspiracy theory on the alliance between corrupted politicians/government and the profit-driven pharmaceutical industry. Many of these posts in fact provide statistics, test results and even seemingly tangible evidence to back their claims. These messages are so elaborated and appear to be trustworthy that even most educated individuals can easily fall prey.

#7

#10. (Google OMC) Spotlight Winner: Michigan State University (United States)

Katie Compo, Zihao Gao and Dakshaini (Daki) Ravinder, pictured with their professor Young Anna Argyris, from Michigan State University. From the team: “We combined our interests in analytics, social media and nonprofits to take up the challenge. Our team brought in fresh perspectives and experience from our jobs and course work to help the client set realistic, achievable and measurable goals, and we went the extra mile to help strategize a plan the client could follow even after the challenge ended.”

unnamed_edited.jpg

What was your favorite part of OMC?

 

The freedom to experiment with ads in a safe environment coupled with the opportunity to work directly with a non-profit partner to improve their marketing efforts. We not only learned valuable marketing skills, but we were a part of teaching our non-profit partner to carry forth our digital marketing plans once our time together came to a close.

What lessons did you take from your experience?

 

The top three things that we learned were:

  • 1. How Google Ads works

  • 2. Specific keywords work better than general keywords

  • 3. The importance of having a user-friendly website

What do you plan to do next with your digital marketing skills?

 

All three of us plan to use our marketing skills in our careers. Katie plans to use her skills to market her photography business, Daki has been using these skills in her job as a Marketing Communications Specialist in higher education, and Zihao plans to use his skills to join a professional sport team’s marketing department.

What advice would you give to other students starting OMC?

Don’t be afraid to stop or adjust an ad set that isn’t performing well. You learn through experimentation and making constant improvements to your ads. Compare, contrast and track your changes so you can take away lessons about what performs most effectively.

#8

#11. Professor Increases HPV Vaccination Rates for Underserved Populations

 

Social media is a realm of free speech and differing opinions. A group of social media users are using this platform to discuss what they believe to be the potential dangers of vaccinating children for HPV (human papillomavirus ), the most common sexually transmitted infection.

Young Anna Argyris, Ph.D., assistant professor of Media and Information, has been selected as the 2019 recipient of the Diversity Research Network Program Launch Awards Program (DRN-LAP) for researching the influence of these messages on mothers deciding whether or not to vaccinate their children. 

Young.png

Social Media and Influence

 

The goal, Argyris said, of her research is to improve HPV vaccination rates among marginalized women. While the professor is one of communication and not of medicine, she saw a niche she could fill in helping to improve these rates. In fact, Argyris’ original research pertained to fake news and misinformation. What she found was that health misinformation had a big impact on our lives.

“The virus is one of the major causes of death, especially among underserved women. It’s definitely an important issue,” said Argyris. “We’ve noticed that anti-vaccination messages play a significant role in suppressing the vaccination behaviors among these groups of women.”

By crawling Instagram and Twitter with a team of students, the researcher noticed those in favor of anti-vaccination used different techniques on platforms to increase vaccine hesitancy among underserved populations. 

Assistant Professor Young Anna Argyris, 2019 Diversity Research Network Program Launch Awards Program Recipient, Investigates Anti-Vaxx Social Media Messaging Phenomenon

A survey was also used to find that mothers’ interactions with the propaganda and influencers negatively influenced their opinions of immunizing their children. More than 90% of decisions made related to vaccinating children were made by mothers, in accordance to the team’s findings. She also stated that more than 90% of US mothers of adolescents are on Facebook or social media accounts. This allows them to be exposed to this messaging.

“They even trust these influencers more than their own doctors. How these influencers talk about these vaccines will influence important decisions on whether these mothers vaccinate their children or not,” said Argyris. 

Awarding Meaningful Change

 

For her research, Argyris received $4980 from DRN-LAP in order to help support her in making any strides she needs to succeed.  The award, given by the Office for Inclusion and Intercultural Initiatives, is “intended to aid scholars in launching new research through pilot study, creative projects of scholarly merit or the enhancement of a measure or technique.” 

Arygyris stated that she would be using the funds to help further this line of research and continue looking into this anti-vaccination social media wave. Plans for the future also include laying groundwork for a program that can filter out misinformation and push down anti-vaccination propaganda and spam.

#9

#12. Faculty Feature of the Month by Health & Risk Communication Center

Our November Faculty Feature of the Month is Dr. Young Anna Argyris. A professor in the Department of Media and Information since 2014, Dr. Argyris’s research focuses on the use of social media to facilitate individual and collective decision-making processes through creating the viral dissemination of content to and exerting social influence on the audiences. Most recently, this work has centered on visual congruence-induced social influence on social media sites for exchanging multimodal content, and the dissemination of health misinformation and its impact on preventive behaviors. To conduct these studies, Dr. Argyris applies a blend of self-reported methods (population-based surveys and in-vivo experiments), large-scale longitudinal data collection and analyses, and machine-learning based automatic classifications.

Dr. Argyris holds her PhD in Management Information Systems from the Sauder School of Business, University of British Columbia. Prior to coming to MSU, she was an assistant professor at the Gabelli School of Business, Fordham University and a visiting scholar at Carroll School of Management, Boston College.

Screen Shot 2020-11-16 at 2.26.18 AM.png

Dr. Young Anna Argyris, Assistant Professor in the Department of Media and Information, is this month's feature of HRCC. Her research focuses on the organizational use of social media to connect disparate parties, using information technology as a decision-making aid.

Her Recent Publications:

 

  • Argyris, Y. A., Muqaddam, A. and Miller, S. (2020) "The Effects of the Visual Presentation of an Influencer’s Extroversion on Perceived Credibility and Purchase Intentions—Moderated by Personality Matching with the Audience," Journal of Retailing and Consumer Services.

  • Argyris, Y. A., Wang, Z., Kim, Y., & Yin, Z. (2020). The effects of visual congruence on increasing consumers’ brand engagement: An empirical investigation of influencer marketing on instagram using deep-learning algorithms for automatic image classification. Computers in Human Behavior, doi:10.1016/j.chb.2020.106443

  • Argyris, Y. A., Wang, Y. and Muqaddam, A. (2020) "Role of Culture in Engaging Consumers in Organizational Social Media Posts," Journal of Organizational Computing and Electronic Commerce, doi:10.1080/10919392.2020.1823177.

  • Argyris, Y. A., Muqaddam, A. and Liang, Y. (2019) “The Role of Flow in Dissemination of Recommendations for Hedonic Products in User-Generated Review Websites,” International Journal of Human Computer Interaction. DOI: 10.1080/10447318.2019.1631543

  • Yim, D., Khuntia, J., & Argyris, Y. A. (2018). User Behaviors and Knowledge Exchange in Health Infomediary. In Handbook of Research on Emerging Perspectives on Healthcare Information Systems and Informatics (pp. 213-233). IGI Global.

  • Wang, Z., Yin, Z. and Argyris, Y. A. "Detecting Medical Misinformation on Social Media Using Multimodal Deep Learning,” is accepted to the IEEE Journal of Biomedical and Health Informatics, https://doi.org/10.1109/JBHI.2020.3037027 (Impact factor >5).

This study centers on the development of a deep learning-based multimodal detector for Instagram anti-vaccine messages (images, texts, and hashtags). To evaluate the proposed model’s performance, a real-world social media dataset that consists of more than 30,000 samples was collected from Instagram between January 2016 and October 2019. Our 30 experiment results demonstrate that the final network achieves above 97% testing accuracy and outperforms other relevant models, demonstrating that it can detect a large amount of antivaccine messages posted daily on social media.

#10
bottom of page