File this in the category of too good to be true: “A new research program by a Cornell computer scientist, in collaboration with colleagues at the University of Pittsburgh and University of Utah, aims to teach computers to scan through text and sort opinion from fact.”
That will revolutionize the pundit business. I’m thinking lots of people will be surprised to discover just how much they read is opinion, not fact. And not just on op ed pages, either. But I’m not hopeful that the tack the researchers are taking will be able to measure anything beside how well a writer disguises his opinion as fact:
The new research will use machine-learning algorithms to give computers examples of text expressing both fact and opinion and teach them to tell the difference. A simplified example might be to look for phrases like “according to” or “it is believed.” Ironically, Cardie said, one of the phrases most likely to indicate opinion is “It is a fact that …”The work also will seek to determine the sources of information cited by a writer. “We’re making sure that any information is tagged with a confidence. If it’s low confidence, it’s not useful information,” Cardie added.
So it’s not like they are actually going to check the writing against facts; they are just going to look at how the information is presented, which means that ironically if you present opinion as fact the programs will take your word for it and flag it as fact. If you are careful and present your opinions as such, then the program will pick up on that and flag it as opinion.
In other words, the researchers are writing a program that uses the writer’s opinion to sort opinion from fact in the writers work. Won’t that be a great help. No word on a computerized sarcasm detector.
