Home.md

@@ -45,9 +45,10 @@ However for adoption of your application- or domain-level metadata and semantics
 - cite the EM-Glossary in appropriate places of your output or publication. 
 
 ### Metadata schemas
-Application level metadata is often collected and validated with metadata schemas. Different formats may be used for this, such as xml or JSON-Schema. We here highlight the latter as a very commonly used way of defining schemas, but envision similar implementations also in other formats or serialisation. 
+Application level metadata is often collected and validated with metadata schemas. Different formats may be used for this, such as XML or JSON-Schema. We here highlight the latter as a very commonly used way of defining schemas, but envision similar implementations also in other formats or serialisation. In JSON schema fields like "external reference" that could carry an emg.owl class label and "identifier" carrying the related class IRIs is encouraged. For linked data schmemas such as YARRRML or JSON-Schema emg.owl base URI should be intergrated as the appropriate context.
 
 ### Semantic file formats
+In semantic file formats, such as HDF5 based containers or NeXus we propose to introduce cross references as doc-strings or separate fields associated with key that carry emg.owl class IRIs. Stronger integration might be achieved through semantic import and alignment of related ontologies. 
 
 ### Semantic artefacts (RDF, RDFs, OWL)
 - import of emg.owl
@@ -57,6 +58,8 @@ Application level metadata is often collected and validated with metadata schema
 - crosswalk with e.g. sssom
 - keep emg IRI
 
+- import with definition (to ensure backwards compatability after new releases of emg.owl).
+
 How to adapt and use terminology from emg.owl in your application level artefact metadata 
 - how to technically adopt
 - how we expect to be acknowledged