ETA's object model
ETA works with both documents and . Internally, these are represented via the ETA Document Model and the ETA Network Model. Within ETA, Document Processing creates the ETA Document Model and Network Creation creates ETA networks from ETA documents.
A generic example of the object structure across documents and networks is shown below.
The object hierarchy
- Community Node - optional
- Resolved Node - optional (includes resolved entities and email threads)
- Simple Node
- Entity - optional
- Reference (‘text reference’)
Objects can be associated via links and connections, as follows:
- between network nodes (that is, between nodes, resolved nodes and communities)
- between document entities
- between network nodes and document entities and text references
- between document text references
- The basic object in a ETA document is the text reference. Depending on the output table settings in the Document Processing configuration a text reference type (for example, Person) may be resolved within the document to become an entity - or otherwise created as an entity, for example via a Document Processing Script.
- The base object of the network is the ‘Simple Node’ which can be created from either an entity or a text reference within the document model via the Network Creation configuration.
- Network entity resolution (that is, cross-document entity resolution) is optional and applies to network nodes.
- Community detection is also optional and applies to resolved nodes.
- Resolved and community nodes are created by clustering nodes at the level below.