AI model detects people's attitudes towards vaccines from their social media posts

People’s attitudes towards vaccines can now be detected from their social media posts with a smart AI model, developed by researchers at the University of Warwick.

An AI-based model can analyze social media posts and determine the author’s attitude towards vaccines, by being ‘trained’ to recognize that attitude from a small number of sample tweets.

As a simple example, if a post contains mention of distrust of health care institutions, fear of needles, or something related to a known conspiracy theory, the model can recognize that the person who wrote it may have negative feelings about vaccinations.

The research, funded by UK Research and Innovation (UKRI), will be presented today (12 July) at the North American Association of Computational Linguistics Annual Conference 2022.

It is led by Professor Yulan He from the University’s Department of Computer Science, supported by a 5-year Turing AI Fellowship funded by the EPSRC.

Professor He and his colleagues at the University of Warwick have used a dataset of 1.9 million tweets in English, posted from February to April 2021, to develop the Vaccine Attitude Detection Model (VADet).

VADet first analyzed the stream of tweets about the COVID-19 vaccine, studying the elements and growing context associated with the ongoing vaccination debate. Then, the model gradually narrows its analysis by looking at patterns that characterize user attention and attitudes.

VADet looks for statistical patterns in words related to different topics or positions. It builds on a large-scale language model that has been pre-trained on a large number of texts from English books and Wikipedia and has acquired some linguistic knowledge. Then they are trained to use vaccine-related tweets so that they understand what topics have been discussed in those tweets.

A small number of those tweets were then manually labeled by the researchers with information about users’ attitudes to the topics covered in the vaccine-related tweets. VADet can utilize a small number of labeled tweets to distinguish semantic information relating to the establishment and topic from the remaining unlabeled tweets.

The AI ​​model then organizes the tweets into groups of similar aspects, forming a geometric pattern that visually shows how a particular point of view on vaccination (pro-vaccination, anti-vaccination, or neutral) can be associated with specific characteristics or references that can be detected in media posts. social. .

This model has the potential to be used to provide insight into why people have negative attitudes about vaccination, information that governments and health organizations can use to better design targeted messages to convince the general public about vaccination.

The COVID pandemic has intensified the use of social media. People express their attitudes towards matters related to public health, including COVID-19 vaccinations. We’ve shown that it’s possible to monitor social media traffic, detect vaccine attitudes, and group tweets into groups that address similar aspects. Such real-time monitoring of public attitudes can help health care organizations and government agencies resolve vaccine doubts and combat vaccine misinformation in a timely manner.”

Professor Yulan He of Warwick’s Department of Computer Science and AI Acceleration at The Alan Turing Institute

The key to the breakthrough lies in a specially developed algorithm, which has two important capabilities. First, it can leverage large-scale social media data on vaccinations to auto-detect topics. This is done by inserting topic layers into existing pre-trained language models.

Second, the algorithm could be adapted to a small set of social media posts labeled with vaccine attitudes to automatically detect topic-specific patterns and attitude-related attitudes. “This so-called adaptive self-repair capability has not previously been explored for vaccine attitude detection,” said Lixing Zhu, a PhD student in Warwick’s Department of Computer Science who implemented the VADet model.

Professor He added: “WHO identified vaccine indecision as one of the top ten health topics in the world in 2019. By automatically detecting vaccine attitudes from social media, our solution has the potential to enable more timely interventions to address vaccination concerns. “

#model #detects #peoples #attitudes #vaccines #social #media #posts