Re: Introductions ...

Hi Aitor,

We can distinguish use cases that serve to explore technical challenges, such as machine learning for knowledge graphs or abductive reasoning, from use cases that explore commercially interesting application areas. I’ve described a roadmap for developing demos that explore technical challenges at:

 https://github.com/w3c/cogai/blob/master/demos/README.md <https://github.com/w3c/cogai/blob/master/demos/README.md>

You will see some pointers to the literature in respect to machine learning under the heading “Learning from examples”.

What kinds of technical challenges would your water pump demo address?

The business oriented demos would need to come after the demos that address the technical challenges that the business demos depend upon. If you have a business opportunity you want to explore, it would be interesting to talk about what technical challenges it entails, and to use this to reprioritise the roadmap for the technical demos. For instance, you might decide that work on demos for certain kinds of reasoning is more important than work on machine learning demos.

A related suggestion could be to focus on developing a tutorial that shows how to extend declarative and procedural knowledge to address a new requirement for a given application. This would start by explaining the initial use case and how it is addressed by a modest sized graph and ruleset, and then talk about an extension to the use case and show how that can be accommodated by changes to the graph and rules.

Happy to chat further.

Kind regards,
Dave

> On 4 Mar 2020, at 08:23, Aitor Corchero <aitorcorchero@gmail.com> wrote:
> 
> Hi Dave, 
> 
> As a use case, maybe it could be interesting to have some use cases to control some pumps in the water distributions for example. We can apply semantic reasoning combines with other AI techniques to make efficient the work of a pump or even a network. The counterpart are the data, that normally are with some noise and requires for data preparation for the analysis... 
> 
> So and form my side, I propose to start with simple examples as you propose and in second iterations goes to more complex and real scenarios. 
> 
> BTW, How is the literature review you asked for? How we can contribute? 

Dave Raggett <dsr@w3.org> http://www.w3.org/People/Raggett
W3C Data Activity Lead & W3C champion for the Web of things