A trio of math and finance professors has created a computerized program to test whether regulators listen to the public when crafting new regulations. The simple answer: Yes.
How they reach that conclusion is far less straightforward. The program, which the professors dubbed “RegRank” in a January paper, uses an automated “machine-learning” method—geekspeak for artificial intelligence—to search through thousands of comment letters from the public and the regulations themselves.
The goal of the system is to discover patterns of words in the documents to determine whether regulators paid attention to the comments, according to the professors, who include Andrei Kirilenko, former Commodity Futures Trading Commission chief economist, Shawn Mankad of the University of Maryland, and George Michailidis of the University of Michigan.
Read the full article [wsj.com]