Grakn for Smart-home Activity Recognition



I am trying to find out that Grakn is suitable for my task or not? I have few Sensors that can send data in real time. I want to represent smart-home environment using Grakn. And identify activities performed by user from raw sensor data and Grakn Scheme based inference. Generally, Ontologies are mostly preferred for this task but, I want to move away from this old practice.


Hi Keyur,

I imagine the data generated by these sensors are rather granular, heterogeneous and complex in their relationships. If that’s the case, then yes, Grakn would be the right database.

With regards to inference and taking advantage of the automated reasoning capability of Grakn, you’d really need to get hands-on to experiment and thereof evaluate. I suggest that you start off from a blog post on defining a basic schema for Grakn and then move on to the docs to model your own experimental knowledge graph.

If you need any assistance along the way, please post your questions here or in the #graql channel on the community slack. The team and the community is quite active in attending to those questions :slight_smile: