Paper Accepted: RAS Special Issue!

Post date: Mar 26, 2019

Figure 16 from our paper: When designing or selecting a knowledge representation, researchers should take into consideration as many of these concepts as possible; this however is a very challenging problem, as there is no fundamental way to integrate these components together into one representation.

My survey paper entitled A Survey of Knowledge Representation in Service Robotics has been accepted in the Robots and Autonomous Systems (RAS) Journal as part of their Special Issue on Semantic Policy and Action Representations for Autonomous Robots. This special issue came about from a workshop series of the same name (abbreviated as SPAR).

In my paper, I give an overview of what knowledge representation means for us researchers in robotics and a survey on representations and models that have been used for creating an effective representation for robots.

It's important to consider how knowledge is represented for a robot to:

1) make learning more efficient,

2) understand the intentions of robots in a way we can easily interpret, and

3) ground knowledge in an easier way.

You can take a look at my paper through this link.


Do you agree with our definition of a knowledge representation?

What sort of standards or ideas can we agree upon? What are some other aspects that should also be considered?

Let's have a discussion!