MIT grad student develops internet lie detector using natural language processing

Nov 22, 2011
0

According to the Nieman Journalism Lab, a graduate student at the Massachusetts Institute of Technology is developing a way to check for lying in political writing on the internet as easily as you check for spelling errors, using novel natural language processing techniques. NLP, for those who aren't familiar with the computer science term, is concerned with the interactions between computers and natural languages, and is essentially a method of human–computer interaction (Apple's Siri on the iPhone 4S is an app utilizing natural language processing).

Partnering up with PolitiFact, MIT media lab graduate student Dan Schultz hopes and expects to “bridge the gap between the corpus of facts and the actual media consumption experience.” Basically, if you're full of BS, or not. The project is utilizing fact-verifying natural language processing techniques, against the information in PolitiFact's API. In English, it means that "it's not able to tell a lie from the truth on its own, but rather it does so by pulling in data on phrases that are in a system." Schultz has plans to open-source it next year, when the project is successfully completed.

If all goes to plan, Schultz’s work could gradually end up being integrated into software that can scan and check websites such as Snopes, allowing internet users a more potent way to debunk claims that so often float around on the Internet as facts, instead of fiction. The MIT graduate student states that “I’m very interested in looking at ways to trigger people’s critical abilities so they think a little bit harder about what they’re reading…before adopting it into their worldview.” Let's hope that this project gets the time and attention it deserves.

[via The Next Web]


Must Read Bits & Bytes