We only have 3 sprints left. Therefore, we need to hurry and make sure that we have a first final draft of our ontology, so that we may begin building certain NLP, machine learning modules.
Include all items of final ontology into the protégé: objects, data science modules, etc.
Understand the objects that need coding: We can decide on what things we actually build in our next team meeting
Communicate new ideas on a regular basis: Add comments to new fields, write in the data science chat, or otherwise communicate (e.g. meetings)
The main goal of the project is to build an ontology, not to build models or pipelines. Therefore, we are 100% free to imagine anything we might want to include in such a project. Let your imagintion flow. If it turns out it is a bad idea, we can discuss resolve them in Protégé.
We only have 3 sprints left. Therefore, we need to hurry and make sure that we have a first final draft of our ontology, so that we may begin building certain NLP, machine learning modules.
The main goal of the project is to build an ontology, not to build models or pipelines. Therefore, we are 100% free to imagine anything we might want to include in such a project. Let your imagintion flow. If it turns out it is a bad idea, we can discuss resolve them in Protégé.
Good luck!