Research & Innovation

This Stevens App Detects Lies and Deception in Text, Social Media

A new mobile app developed at Stevens tells you, in an instant, who's telling the truth.

Jaasuz, a revolutionary technology invented at Stevens by two faculty members and perfected with the assistance of their graduate students, is available as a Google Play Android application following several years of development. It may be the first-ever mobile app to truly tell right from wrong. (A web version of the app is also available online.)

The app reads text and quickly sifts through it for dozens of different clues about truth (or intent to deceive), as well as gender, drawing on historical patterns extracted from hundreds of confirmed online hoaxes and half a million known-gender emails. Potential applications for the fraud-detecting technology might include uses in the insurance, law enforcement, cybersecurity and legal professions, among others.

"We have refined the app and the algorithms a great deal recently," says Rajarathnam "Mouli" Chandramouli,  Thomas E. Hattrick Chair Professor of Information Systems at Stevens, who devised the technology in collaboration with Associate Professor of Electrical and Computer Engineering K.P. Subbalakshmi. "It now can do very accurate analyses of texts, based on very few words.

"We think enterprises will be extremely interested in this capability, particularly given the number of cyber attacks that originate as tweets, text messages or social media posts."

Additional Jaasuz tools available online can also detect malicious or sexual intent, same-authorship and plagiarism, adds Chandramouli.

The tool is already receiving national media attention from outlets including FOX News.