IRI patterns on the mapping tool and Variables in the DTS

[Wenjie225]

For UM's data (breast cancer case), we could have the same IRI pattern (source) for multiple nodes depending on the variables in DTS. To some extent, it shows the difficulty to link some variables from DTS to current classes in SPHN schema.

Example: In discharge summary (UM's data), we only have variable names called "ALLERGIES_TYPE" and "ALLERGIES_DATE" for allergy information, while in the SPHN schema, there are classes such as "Allergy", "Allergen", and "Allergy Episode". We added a data node called "Allergy", linked it to the "ALLERGIES_TYPE" at first. But when we added a connection "has allergen" and tried to link it to a new data node called "Allergy type", we needed to type in the same IRI pattern as in the node "Allergy". Then we created another connection between "Allergy type" and a new node "Food allergy" to show the exact allergy type, with the same IRI pattern as in "Allergy".

Finally when we tried to complete the mapping, the system showed an error: "Multiple sources on Value field found"

So in this case, will there be any effect on the further knowledge graph when we use the same IRI pattern (source) for several data nodes? If so, we may have to consider solving this by either adjusting variables in DTS to make it more compatible with the current ontology, or extending/changing classes.