To successfully compete on Jeopardy! Watson had to analyze each question to determine exactly what was being asked, analyze the available content to extract precise answers, and quickly compute its level of confidence in the answer based on supporting and refuting information found.
In its process of analyzing a question and determining the best answer, Watson applies advanced natural language processing, information retrieval, knowledge representation and reasoning, and machine-learning technologies to open-domain question answering. At its core, Watson is built on IBM DeepQA technology for hypothesis generation, massive evidence gathering, analysis and scoring. It has been loaded with millions of documents, including dictionaries, encyclopedias, taxonomies, religious texts, novels, plays and other reference material that it can use to build its knowledge.
Unlike search engines, which can understand basic questions and return answers in the form of a list of relevant documents, Watson can understand questions posed in natural language and returns precise answers that directly answer the question. Rather than relying on a single algorithm, Watson uses hundreds of algorithms to find candidate answers, score those answers, gather additional supporting evidence for each candidate answer, and deeply evaluate that evidence using complex natural language processing. The more algorithms that independently arrive at the same answer, the higher the confidence level. Watson weighs all of the evidence for each candidate answer to identify the best answer and determine a confidence for that answer. If the confidence level is high enough, Watson buzzes in and answers the question.
The Watson technology has the potential to evolve for commercial use, transforming a number of industries. For example, in healthcare, it could be deployed as an online tool to assist medical professionals in formulating a diagnosis. A doctor could input a set of symptoms and medical history and receive a possible diagnosis to help the doctor arrive at a final diagnosis and treatment plan. Retailers could employ the technology to help shoppers find the exact item they are looking for. For travelers, it could help map out the most viable vacation options or transportation routes.
“The challenge is to build a system that, unlike systems before it, can rival the human mind’s ability to determine precise answers to natural language questions and to compute accurate confidences in the answers. This confidence processing ability is key. It greatly distinguishes the IBM approach from conventional search and is critical to implementing useful business applications of Question Answering.”
“IBM Developing Computing System to Challenge Humans on America’s Favorite Quiz Show, Jeopardy!”April 27, 2009