Levels
Maturity Level 1¶
- Business terms for EKG use cases are captured and mapped to an ontology (possibly as simple labels)
Maturity Level 2¶
- Relevant terms (for EKG use cases) are associated with the use case independently of their corresponding ontologies
Maturity Level 3¶
- Existing glossaries within the organization, within the scope of supported use cases, are mapped to ontologies and imported into the EKG
- Terms are grouped into vocabularies for reuse in different communities
- Vocabularies include local textual definitions as well as being mapped to ontologies
- Community level vocabularies able to import common vocabularies reusable in many communities: allow for multiple levels from local to global
- Support for synonyms and abbreviations within vocabularies
- Vocabularies use general principles of modularity
- Ontologies presented to different communities in their own language
- Natural language used for definitions links to other terms used
- Terms are searchable via the knowledge graph interface
- Generate vocabularies from existing data models or ontologies
- Ontology logic is presented to business users as natural language
Maturity Level 4¶
- Ontology logic is presented to business as natural language
- EKG is the authoritative source for all terms, scoped by community, context and use case
- The governance processes for all new terms (and changes) are managed directly within the EKG
- The results of ontology inferencing are presented in business natural language
- Use of ontology logic to validate use of vocabulary terms
- Management of homonyms and disambiguation of terms---through being mapped to different concepts
- Use of business terms in natural language queries
- Techniques for disambiguation e.g. allowing business users to select the intended meaning
- Use of textual analytics to suggest modifications to vocabulary/ontology (lemmatization)
- Use of vocabularies to match internal databases or external data sources for linked data,
and establish
sameAslinks
Maturity Level 5¶
- Terminology is used to support natural language processing (NLP) of unstructured data for the EKG
- Use of reference data from the EKG to support entity resolution (in EKG context or elsewhere)