Re: Introductions ...

Hi Aitor,

Welcome to the Cognitive AI CG!  Given your experience with knowledge graphs, do you have any suggestions in respect to use cases and data sets for a demo to explore cognitive approaches to learning knowledge graphs from a sequence of examples?

This would start from a set of examples of relationships and properties, and apply inductive reasoning to create an ontology that provides a good fit to the examples, taking their statistics into account. There could be examples that don’t fit, perhaps due to an insufficiency of data, or alternatively, due to the presence of noise in the examples. The aim is to show effective learning from relatively small sets of examples, and to contrast cognitive approaches with deep learning.

One potential use case is to devise a taxonomic classification that accounts for descriptions of particular animals, e.g. how many legs, presence of fur or feathers, skeletal characteristics and so forth. I am sure that there are many others, e.g. predicting someone’s likely interest in a given product based upon their past behaviour, and that of others like them. The challenge is to find data sets we can use in public demos.

Best regards,
Dave
 
> On 23 Feb 2020, at 10:10, Aitor Corchero Rodriguez <aitor.corchero@eurecat.org> wrote:
> 
> Hi all,
> 
> I'm Aitor Corchero a researcher at Eurecat [1]. My main experience is on semantic web technology specially applied to water domain. In our solutions, we merge Semantic Web with AI technologies. We applied some similar approach on Cognitive Science some years ago by merging knowledge-graph with Agent-Based modelling to manage water resources.
> 
> So, we are happy to contribute to the group!
> 
> [1]: https://eurecat.org/
> 
> BR,
> 
> Aitor Corchero

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

Received on Sunday, 23 February 2020 11:11:58 UTC