Uncover Insights on the Public's Opinion of Measles from Twitter


Introduction

One of the greatest gifts given to humanity by science has been the improvement in quality of life. Science has brought us medicine, vaccines, and medical tests that can protect and improve the health of most people. However, we are at a point as a society where the access to information at our fingertips on the Internet has opened a door for other messages to enter: those with mis- and disinformation. Due partly to this mis- and disinformation, a mistrust of vaccines has been sown into certain people in our society and there has been a resurgence of diseases that were once eliminated in the United States, including polio and measles. Consequently, it has become important to allow public health agencies and doctors to quickly be alerted to emerging mis- and disinformation trends, so that vaccine messaging and communications can be clarified. This type of application has the potential of supporting public health interventions so that innocent people, such as children, do not end up plagued by diseases that were easily preventable with a vaccine.


To get a peek into the mind of the general public, professionals can turn to social media. Various forms of social media document the thoughts and values of our society in real time, as well as how these trend over time.


Overview/Background, Methods Used

Dr. Jingcheng Du, Director of NLP Research at Melax Tech, in collaboration with a talented team of researchers from the University of Texas and Texas A&M, harnessed the power of NLP, deep learning, and multi-task learning to gain insights about the public’s opinion of measles and the vaccine for it. The team analyzed a total of 1,917,032 tweets posted between Jan. 18, 2007, and Apr. 23,2019 classified them according to the categories Attitude towards Vaccination, Emotion Expressed, and Type of Message. Melax Tech’s technology was used to compare eight different machine and deep learning algorithms.


Figure 1: Architecture of the MT-CNN Model


Applications to Health

These types of tools can allow researchers to analyze both retrospective and real-time data from social media platforms, resulting in improved vaccination promotion and health communication programs. They provide the opportunity to adapt current programs in real time to counter emerging mis- and disinformation. While cognitive dissonance is not something that can be easily remedied, by utilizing data from the public through social media, it may be possible to reroute it so that fact-checked and peer-reviewed information are more likely to make an impact on the public’s opinion of life-saving vaccines and other health interventions.


The technology and methodology used in this work can be utilized by public health professionals and their communications and marketing teams to understand the public’s opinion on health, their opinion trends over time, and what drives them to make decisions that impact the overall health of a society. Melax Tech’s CLAMP contains award-winning algorithms and other technology that can be applied to different contexts across healthcare in hospitals and clinics as well as be utilized by the pharmaceutical and insurance industries.


A full description of our methodology and results is given in “Understanding Public Perceptions of Measles from Twitter Using Multi-Task Convolutional Neural Networks” by Samuel Wang, Jingcheng Du, Lu Tang, Cui Tao. Studies in Health Technology and Informatics, Volume 290: MEDINFO 2021: One World, One Health – Global Partnership for Digital Innovation, pages 607 – 611, https://ebooks.iospress.nl/doi/10.3233/SHTI220149


Does your organization want to utilize the massive data sets that are constantly being developed with every tweet, hashtag, and post? Melax Tech can help to harness that information to gain insights that will benefit your organization. You can request a demo by clicking on the button below.