Abstract
•We introduce the idea of human processed AI based on knowledge learned by humans.•Note information provided by human experts is typically linguistically expressed.•Look at properties of ordinal scale to model linguistically expressed information.•Discuss modeling of information about an uncertain variable using an ordinal scale.•We look at problem of multi-source fusion in this ordinal environment.
In human processed AI, HP-AI, we build our AI systems based on knowledge learned by human experts rather then that learned by artificial neural networks such as in the case of deep learning. The information provided by these human experts is typically linguistically expressed. In support of HP-AI we look at the properties of an ordinal scale, S, needed to model linguistically expressed quantitative information. Since fuzzy measures provide a very general structure for modeling uncertainty we look at ordinal fuzzy measures. We look at the Sugeno integral based on this ordinal S scale. We discuss the modeling of information about an uncertain variable using an ordinal scale. We look at the problem of multi-source in this ordinal environment.