Watson is very unique in that it is fundamentally reasoning over unstructured information. Traditionally, when you think of finding information you think of using keyword matches, or parts of speech, or a variety of different techniques. Watson is a sum of all techniques.
What makes Watson unique is that instead of having one particular language or one particular technique, or one particular technology that we would use to find information, we pick the best techniques that working together will provide the right answer. So it is very unique and very novel in this way, and its many languages and its many technologies, and many algorithms all work together in a sort of a symphony to find information.
If I think of a Watson use case, or ‘what is a use case Watson is good at’, what I think of are tasks that a human takes an enormous time to do. If people today are taking an inordinate amount of time to look for information using various techniques and thinking through a problem, that’s a good use case. A good example is a doctor looking at specific information, a patient record or existing guidelines or any information, and trying to figure out what the best treatment is for a patient. A human is spending tremendous time and effort doing cognitive reasoning over the information to give an answer, which is what makes it a good use case.
Again, when someone has to spend a tremendous amount of time looking for information, and Watson can augment their ability or replace their need to do that, it’s a great use case for Watson.
See on www.ibmdw.net