How AI could help simplify politics

A visitor watches an AI (Artificial Intelligence) sign on an animated screen at the Mobile World Congress (MWC), the telecom industry's biggest annual gathering, in Barcelona. CR quizzed a handful of popular, general-purpose AI chatbots to see if their advice on health and safety topics matched that of our experts. (Photo by JOSEP LAGO/AFP via Getty Images)
A visitor watches an AI (Artificial Intelligence) sign on an animated screen at the Mobile World Congress (MWC), the telecom industry's biggest annual gathering, in Barcelona. (Photo by Josep LAGO / AFP/Getty Images)

WEST LAFAYETTE, Ind. (WIBC) — If you’ve ever read legislation and didn’t think it made sense, a Purdue University professor says there is a method based in artificial intelligence that could help.

It’s called natural language processing (NLP).

“One of the things that NLP can help with is in processing that language and providing better options for the public. A lot of the proposed legislation is written in legalese or a language that is not directly accessible to the public,” said Dan Goldwasser, an Associate Professor in the Department of Computer Science at Purdue University.

Goldwasser says NLP can summarize legal documents that are written in a way that is hard to understand for the average person.

“Having a system that can help decipher what the intended policy actually means, what complications it will have on different segments of the public and professions, can help people develop more informed views,” said Goldwasser.

He does acknowledge, however, that there are concerns with NLP.

“These systems are not trained through dedicated teams of experts carefully picking data, but rather they are trained over massive amounts of text that are publicly available. These texts capture many good things like scientific knowledge, common sense inferences, and so on. But they also capture the kind of biases that exist unfortunately in our society,” said Goldwasser.

Goldwasser says those inherit biases about certain groups can lead to skewed or inaccurate results.