Obviously, when it comes to combining natural language queries with advanced text analytics, the IBM Watson supercomputer, as so evidenced on the "Jeopardy!" game show, represents state-of-the-art computing.
But perhaps even more interesting is where IBM intends to take Watson next. Watson relies heavily on the open source Unstructured Information Management Architecture (UIMA) specification that is managed by the Apache Software Foundation. In the not-too-distant future, UIMA will be extended to include support for both audio and video, which means in addition to searching text, Watson will soon be able to search through troves of audio and video files, according to Dr. David Ferrucci, an IBM fellow and principal investigator on the Watson/Deep QA Project.
But even more intriguing is the fact that in addition to providing answers, IBM is also working on Watson's ability to shape questions. Initially, most of that work is going to focus on fine-tuning the original query by comparing that question based on previous similar interactions to enable Watson to come up with follow-up questions. But longer term, IBM is also working on what Ferrucci describes as "Active Learning" that will allow Watson to be more proactive about formulating questions based on the data it collects.
It will be a long time before most organizations can afford a Watson supercomputer, but don't be too surprised to see Watson driving some new analytics-as-a-service offerings aimed at various vertical industries. After all, in the age of the cloud, we're learning all the time that you don't actually have to own a machine to make use of it.