Algorithm Reads Millions of Posts To Understand Online Misinformation

Message board forums can be easily parsed provided one spends a significant amount of time understanding the logic within them. In an attempt to better understand how human communicate and think online, University of California Lose Angeles researchers are on a mission to teach computers a way to understand the chronological posts on the sites.

"Our question was, could we devise computational methods to discover an emerging narrative framework underlying internet conversations that was possibly influencing the decision making of many people throughout the country or possibly world?" said UCLA researcher Timothy Tangherlini.

Tech Radar said an algorithm was built by his team to review messages posted in two parenting forums. The messages numbering 1.99 million and posted by 40,056 users most of them mothers, were posted in a period of nine years that ended in 2012. Tangherlini said the anti-vaccine was clearly a good piece for the study so they focused on these posts.

"Tens of thousands of parents were exchanging ideas about child-rearing online and, through those interactions, creating virtual communities where they could share concerns, propose methods to allay those concerns, and share their own experiences," Tangerlini said.

What Is said the researchers discovered that many parents were posting messages about vaccines including their risks, the trustworthiness of institutions and how they can exempt their children from vaccination.

An algorithm is a problem-solving formula much like a computer program that is used widely in information technology. Algorithms make use of keywords to find relevant results.

JMIR Publications said forums are very important in determining the trending topics and issues being discussed by internet users nowadays. Parents, for one, patronize health sites to discuss various healthcare issues that affect their families. Researchers used probabilistic method to determine the topics being discussed by parents on these sites. The algorithms allowed the researchers to sift the underlying stories based on fragmented posts.