After a paper is published, say in CS, who reads the paper? How are the ideas presented there applied to business and industry? How is industry collaboration done?
I'm asking this question per this comment.
I think it depends in large part on what the research is. Some fields lend themselves more readily to real-life application, while others like, say, theoretical physics, are a long way from being used by industry.
It also depends on the researcher. I am a firm believer in the "Last Research Mile" principle (disclosure: the man who wrote that is my advisor). The premise of the Last Research Mile is that a big part of doing research is taking it all the way through to implementation. Doing the early research provides a good start, but lots of important lessons, including research lessons, are learned as the idea is carried through to execution. I'm studying in an Information Systems program, and so my research is in business, computers, and people. While it would be easy to do the basic research and move on to another topic, I think it is important to continue to push and test the idea even in its implementation phases.
As a result of following the Last Research Mile principle, my advisor has created at least one company that sells the product of decades of his research. Many lessons were learned, and many publications resulted because of those lessons throughout the whole process.