As the COVID-19 pandemic grew, the World Health Organization and the United Nations issued a stark warning that the “infodemic” of online rumors and fake news about COVID-19 was hampering public health efforts and causing unnecessary deaths. “Disinformation is costing lives,” the organizations warned. “Without proper trust and accurate information … the virus will continue to thrive.”
In an effort to solve this problem, researchers at the Stevens Institute of Technology are developing a project to discover an Artificial Intelligence (AI)-based application that is capable of detecting "fake news" about COVID-19 and automatically flagging misleading news reports and social media posts, writes the Stevens Institute of Technology in New Jersey, USA.
“During the pandemic, things became extremely polarized,” explained Koduvayur Subbalakshmi, an artificial intelligence expert and professor of electrical and computer engineering at the Stevens Institute.
"We urgently need new tools to help people find information they can trust," she added.
To develop an algorithm capable of detecting misinformation about COVID-19, Subbalakshmi first worked with Stevens graduate students Mingxuan Chen and Xingqiao Chu to collect about 2600 news articles about COVID-19 vaccines, published by 80 different publishers over 15 months.
The team then collected over 24,000 Twitter posts that mentioned indexed news reports and developed a “stance detection” algorithm capable of determining whether a tweet was supportive or opposing of the article in question.
Using their labeled datasets, Stevens’ team trained and tested a new artificial intelligence architecture designed to detect subtle linguistic cues that distinguish real reports from fake news. This is a powerful approach because it doesn’t require the AI system to audit the factual content of a text, or keep track of the evolution of public health messages; instead, the algorithm detects stylistic fingerprints that correspond to credible or untrustworthy texts.
“It is possible to take any written sentence and turn it into a data point – a vector in N-dimensional space – that represents the author’s use of language. Our algorithm examines those data points to decide whether an article is more or less likely to be fake news,” explained Dr. Subbalakshmi.
More bombastic or emotional language, for example, is often associated with false claims, Subbalakshmi explained. Other factors such as the time of publication, the length of an article, and even the number of authors can be used by an AI algorithm, allowing it to determine the credibility of an article. These statistics are provided with their newly curated data. Their basic architecture is able to detect fake news with about 88% accuracy – significantly better than most previous AI tools for detecting fake news.
This is an impressive discovery, especially using data that was collected and analyzed in almost real time.
However, much more work is needed to create tools that are powerful and rigorous enough to be deployed in the real world.
(The full article is published in www.stevens.edu)