Recording from the Software Center Brown Bag seminar on November 8th:
”Setting AI in context: Why data for AI needs context”
The seminar was hosted by theme 4 (‘Customer Data and Ecosystem Driven Development’) and the speaker is Hans-Martin Heyn. Hans-Martin is a Postdoctoral researcher at the Department of Computer Science and Engineering at Chalmers/University of Gothenburg.
For automated driving systems the operational context needs to be known in order to state guarantees on performance and safety. We show, based on semi-structured interviews, that major uncertainties exist in how to clearly define and document the operational context in a diverse and distributed development environment such as the automotive industry. This is problematic especially for systems using machine learning, because data attributes and data quality for training and validation depends on the assumed context of the system. Based on Design Science Research, we propose a Data Quality Assessment and Maintenance Framework, to systematically manage data quality and related requirements.