Microsoft Research Connection's Semantic Computing initiative strives to accelerate research in machine learning and artificial intelligence to develop innovative products and services that use rich, semantic information.
Semantic computing is about exploring complex data from disparate sources and making sense of that data. It is about interacting with data in rich, safe, and semantically meaningful ways to expose the path from data to information, knowledge, and intelligence. Semantics have the potential to unify our approach to data encoding, thereby facilitating access to and use of data, and ultimately helping researchers gain new insights.
To reach these goals, we strive to create research programs that investigate computational intelligence to facilitate open innovation and agile experimentation via semantically-rich, cloud-based services, as well as projects to advance the state of the art in knowledge representation and reasoning under uncertainty at web scale.
To progress from data services to knowledge services we must examine the future of information processing and discovery at web scale and use the power of collective intelligence. To do this, each year Microsoft hosts a series of “mind swaps” and “think tanks,” during which we explore how semantic knowledge could be transformed in a way that is useful for commodity computing and that aids in the creation of knowledge services that are relevant to the research community.