Narrator: The field of artificial intelligence continues to explode, digital assistants, facial recognition systems and self-driving cars are just the tip of the ice burg as computers improve and the systems become more complex.
Narrator: Working on improving AI and how it understands human writing was the aim of research being conducted by Craig Olson, a student of Information Systems at the University of Nebraska Omaha.
Insert Sound Byte 1 “Craig Olson” 0:20-0:35
Narrator: A question-answer system is a computer program that a human can type in a question using natural languages, like how we talk in everyday life, and the system would be able to respond to the inquiry.
Narrator: Currently, most companies that deal with language processing use a form of machine learning, this is where large amounts of data are fed to a computer that then looks for patterns and use probability to give proper responses back. This requires huge amounts of computational power.
Narrator: Craig explains that part of the reason why we humans can understand sentences so easily and computers struggle is because we have a common-sense understanding of the world, we live in. People understand you must have an action before a reaction. Craig looked to bring that type of reasoning to computers.
Insert Sound Byte 2 “Craig Olson” 1:15 – 1:51
Narrator: Not only could these computer systems understand us better but they could also become more resilient to what Craig calls “noisy data” or data that is nonsense. This nonsense data can cause machine learning systems to send out wildly inaccurate information.
Insert Sound Byte 3 “Craig Olson” 2:05 – 2:42
Narrator: Craig thinks as the field continues to rapidly change, systems like the one he developed along with others could come together to make far more advanced systems than the ones we are used to now.
Narrator: For the UNO School of Communication, I’m RC Miller.
MAV Research 