Microsoft creates AI that can read a document and answer questions about it as well as a person - The AI Blog


Microsoft researchers have created technology that uses artificial intelligence to read a document and answer questions about it about as well as a human. It's a major milestone in the push to have search engines such as Bing and intelligent assistants such as Cortana interact with people and provide information in more natural ways, much...

HPE enters natural language question answering fray


Hewlett Packard Enterprise has added natural language question answering to its unstructured data analytics engine. The engine, HPE IDOL, uses machine learning to boost the accuracy of human interactions with computers. The next wave of IT innovation will be powered by artificial intelligence and machine learning. We look at the ways companies can take advantage of it and how to get started. HPE IDOL's natural question answering capability joins a crowded field as Nuance and IBM via Watson all have similar plays on natural language processing.

Microsoft AI reads documents, answers questions better than humans


Some bemoan how present society has devalued reading skills and comprehension, which doesn't bode well for human civilization as a whole. To add insult to injury, it seems that computers may soon become even better than humans at comprehending what they read. The Microsoft Research team in Asia have created an artificial intelligence that can not only read a document, it can even answer questions about said document, scoring higher than humans on that same test.

Answer Selection and Confidence Estimation

AAAI Conferences

We describe BBN's Question Answering work at TREC 2002. We focus on two issues: answer selection and confidence estimation. We found that some simple constraints on the candidate answers can improve a pure IRbased technique for answer selection. We also found that a few simple features derived from the question-answer pairs can be used for effective confidence estimation. Our results also confirmed earlier findings that the World-Wide Web is a very useful resource for answering TREC-style factoid questions.