Relation Features

Relations are associative concepts. In the NLP API you can think of them as a set of relationships between one or more entities and the attributes specific to those relationships. Relations can be simple, involving only 2 entities in a single relationship, or complex, involving multiple relationships, and multiple entities. An example of a simple relation is the qualifier relation. This relation consists of two arguments where one is the entity being described, and the other is the descriptive term. An example of a complex relation is a follow-up relation which models a follow-up recommendation and consists of three different relationships: the reason for the follow-up, the recommended procedure and timeline.

The NLP API supports the following feature relation types:

• Experiencer relation
• Measurement relation
• Medication relation
• Qualifier relation
• Anatomic site relation
• Temporality relation
• Reported Event relation
• Follow-up relation
• Imagelink relation

Extraction of relations must be enabled during processing, see Submitting Documents.