The Mohamed bin Zayed University of Artificial Intelligence in Abu Dhabi has introduced a new tool to help identify persuasion tactics in media reports.
The tool goes by the acronym Frappe – framing, persuasion and propaganda explorer. It was developed by Preslav Nakov, professor and chairman of natural language processing at the university, the world's first higher-learning institution dedicated to AI.
"It was developed together with the European Commission, and a number or journalists and research institutions across Europe," Prof Nakov, who has a long history in the field of disinformation analysis, explained in a video posted to MBZUAI's social media accounts.
"To the policymakers and journalists, it offers the opportunity to study and compare different countries, different languages, different media, and to the general user it offers a very useful media literacy tool."
Frappe allows users to paste the URL link or text from any news article and provides insights as to how those pieces are framed.
It also offers an analysis of the persuasion techniques and provides users with the ability to compare and contrast various persuasion styles used in countries throughout the world, with framing classifications including political, economic, morality, fairness, crime and security.
"You can see a visualisation of 2.5 million articles across 186 countries along with very cool visualisation tools," Lara Hassan, an MBZUAI student, said in the video.
MBZUAI said the Frappe tool has been presented at the Conference of the European Chapter of the Association for Computational Linguistics.
Frappe started as an undergraduate research internship programme several years ago and has since been developed by researchers around the world, MBZUAI said.
The tool's data set is 1,600 articles in English, French, German, Italian, Polish and Russian, though it has the ability to process reports in more than 100 languages.
MBZUAI said researchers hope to expand the training data set and enhance Frappe's accuracy as development continues.
Prof Nakov pointed out the nuance of persuasive language, explaining that although its use need not necessarily be viewed as a negative, it can be abused to misinform, especially with the involvement of AI and fake social media accounts.
"We have seen machine-generated content put online with the purpose to deceive," he said. "We see fully automatic websites for fake news. And I anticipate more fully automatic multimedia content in the future.”
MBZUAI was first announced in 2019, as AI developments continued to generate excitement after years of research in various technology and academic circles. The university opened in 2020.
