Microsoft Research releases Concept Graph to help computers conceptualize in a humanlike fashion


Today, a team of scientists from Microsoft Research Asia, Microsoft’s research lab in Beijing, China, announced the public release of technology called ‘Concept Graph’ designed to help computers conceptualize in a humanlike fashion.

For example, consider a problem in natural language processing. Humans do not find sentences such as “animals other than dogs such as cats” ambiguous, but machine parsing can lead to two possible understandings: “cats are animals” or “cats are dogs.” Common sense tells us that cats cannot be dogs, which renders the second parsing improbable. To make computers understand like humans, we need to make them have knowledge about concepts (e.g., persons and animals) and the ability to conceptualize (e.g., cats are animals).

Microsoft Research built a research project called Probase, which is a big graph of concepts. Probase is built using billions of web pages and years’ worth of search logs. This Microsoft Concept Graph release is built upon Probase.

The public release of the Microsoft Concept Graph and Microsoft Concept Tagging Model are intended to support research on natural language understanding for technologies such as search engines, chatbots and other artificial intelligence systems.

Read more about it here.