A company funded by the U.S. government is using artificial intelligence to detect and delete coronavirus-related misinformation. This marks another instance where technology has been seen as an answer to the outbreak.
Since the pandemic intensified, governments have been relying on tech solutions to handle the burgeoning crisis. Various tech companies have come up with digital tools for tasks, ranging from detecting those with a fever to self-screening for COVID-19 symptoms. Although, the majority of the efforts appear to have focused on developing tracking or AI-powered facial recognition software developed in China that uses data from street cameras to detect people with symptoms, even when they are wearing masks.
This latest tool is another example of how effective AI is becoming in helping to solve real-world problems. As reported by The , the machine learning software is the product of S-3 Research, established by Timothey Mackey, an associate professor of health sciences at University of California, San Diego. It was originally created for finding potential drug dealers on the internet by using a trained classifier to detect illegal opioid ads using Natural Language Processing (NLG) techniques. Now, the AI-powered solution has been recalibrated to detect coronavirus misinformation. Over time, scams related to the pandemic have evolved from vitamin tablets and herbs to ads trying to sell fake testing kits.
How AI Will Find COVID-19 Misinformation
In order to retool for the fight against coronavirus misinformation, Mackey and the team let the software scrape more than 80 million posts from Twitter, Reddit and LinkedIn. Texts from these posts were then analyzed using NLG algorithms for a combination of words suggesting the outbreak and sales. A classifier is being trained to detect posts with the suspected word combos that are linked to shopping sites. This is a challenging task as the features used to train the classifier are not foolproof with scammers often and quickly changing tactics. While Facebook's AI algorithm removed posts with accurate COVID-19 information. Considering how busy social media platforms are at the moment, help from third-party solutions could prove to be useful.
Authentic information is important for managing a public health crisis. Misinformation, such as the their services for free to help those who become targets of fake emails containing malwares. Although, educating s to identify misinformation, coronavirus or otherwise, might be the only real way to solve the issue completely.
Source: The