Burning Question: Is it possible to define data and data relations through speech, with the aid of Natural Language Processing? And what opportunities would there be in that?
Aarhus University, M. Sc. Computer Engineering
The goal of this project, is to be able to turn input in the form of natural language into a computer understandable data definition.
There is already been done some work in the field of generation or extending an ontology, but the current work requires supervision and large collections of domain specific data. The approach applied in this project, tries to reduce these two factors. This would make it more feasible for the common user to extend a common ontology, and use their own extensions in cooperation with extensions made by other parties.
There are several interesting use cases for this. The use case that inspired me for this project, is a user being able to explain to an artificial assistant, what different "things" are, and how the artificial assistant should manipulate these "things".
Hopefully this will enable a faster and more seamless integration of an artificial intelligence, in specifically the tasks the user has a need in, removing the requirement for developers to "think like the user".
UPDATE January 2018: Morten just signed a contract with Stibo Systems and will start working in our software development department with a clear focus on machine learning. Congratulation :-)