|
1 | 1 | # Concept |
2 | 2 |
|
3 | | -For every given Use Case we want to start with capturing the concepts and |
4 | | -terms that the user or "the business" uses or wants to use. |
5 | | - |
6 | | -Most of these concepts and their terms will be pre-defined in all kinds of |
7 | | -vocabularies but for brand-new use cases in a new domain concepts and their |
8 | | -terms will have to be created. |
9 | | - |
10 | | -At the initial stages of a use case, the focus should be on capturing the |
11 | | -language of the users in their domain, which may not necessarily involve |
12 | | -discussing ontologies. |
13 | | -The main goal is to gather requirements and understand the problem context, |
14 | | -as well as the terms and concepts used by the users. |
15 | | -Later in the use case's life cycle, once the problem is well understood, |
16 | | -the relevant ontologies can be mapped to the terms and concepts captured earlier. |
17 | | -This allows for better integration of the use case with the overall EKG ecosystem. |
18 | | - |
19 | | -As the use case evolves and the understanding of the domain becomes clearer, |
20 | | -it may be necessary to adjust the captured concepts and terms to better |
21 | | -reflect the reality of the domain. |
22 | | -, it may be necessary to map the captured concepts and terms to more appropriate |
23 | | -terms in a specific ontology or vocabulary to ensure consistency and interoperability |
24 | | -across the enterprise. |
25 | | -In either case, the important thing is to ensure that the captured concepts and |
26 | | -terms accurately reflect the reality of the domain and the needs of the stakeholders. |
| 3 | +<!--summary-start--> |
| 4 | +_A domain term or idea that links business language to semantic models, |
| 5 | +enabling the Enterprise Knowledge Graph to understand and reason about |
| 6 | +business concepts_ |
| 7 | +<!--summary-end--> |
| 8 | + |
| 9 | +=== "Business & Management Audience" |
| 10 | + |
| 11 | + ## What Is a Concept? |
| 12 | + |
| 13 | + For every given [Use Case](use-case.md), we want to start with |
| 14 | + capturing the concepts and terms that the user or "the business" |
| 15 | + uses or wants to use. |
| 16 | + |
| 17 | + A **Concept** is a way to represent a business idea or term in a |
| 18 | + way that both humans and machines can understand. |
| 19 | + It serves as a bridge between the language your organization uses |
| 20 | + and the formal semantic models (ontologies) that enable the |
| 21 | + Enterprise Knowledge Graph to reason about your business. |
| 22 | + |
| 23 | + ## Why Concepts Matter |
| 24 | + |
| 25 | + Concepts ensure that: |
| 26 | + |
| 27 | + - **Business language is preserved** — We capture terms as the |
| 28 | + business uses them, not as some external standard defines them |
| 29 | + - **Semantic meaning is enabled** — Concepts link to ontologies |
| 30 | + that give them machine-readable meaning |
| 31 | + - **Consistency is maintained** — The same concept can be reused |
| 32 | + across multiple use cases |
| 33 | + - **Interoperability is achieved** — Concepts can map to |
| 34 | + multiple ontologies, enabling integration across systems |
| 35 | + |
| 36 | + !!! tip "Start with business language" |
| 37 | + |
| 38 | + Don't worry about ontologies in the early stages. |
| 39 | + Focus on capturing what the business calls things and what |
| 40 | + those things mean in their context. |
| 41 | + |
| 42 | + ## Concept Vocabulary |
| 43 | + |
| 44 | + Most concepts and their terms will be pre-defined in all kinds of |
| 45 | + vocabularies, but for brand-new use cases in a new domain, |
| 46 | + concepts and their terms will have to be created. |
| 47 | + |
| 48 | + Each Use Case can have its own vocabulary of concepts, but it can |
| 49 | + also inherit or borrow concepts from higher-level or related use |
| 50 | + cases in the [Use Case Tree](use-case-tree.md). |
| 51 | + This enables reuse and consistency across the enterprise. |
| 52 | + |
| 53 | + ## Evolution and Refinement |
| 54 | + |
| 55 | + As the use case evolves and the understanding of the domain |
| 56 | + becomes clearer, it may be necessary to: |
| 57 | + |
| 58 | + - **Adjust concepts** — Better reflect the reality of the domain |
| 59 | + - **Map to ontologies** — Link concepts to more appropriate |
| 60 | + terms in specific ontologies or vocabularies to ensure |
| 61 | + consistency and interoperability across the enterprise |
| 62 | + |
| 63 | + In either case, the important thing is to ensure that the |
| 64 | + captured concepts and terms accurately reflect the reality of the |
| 65 | + domain and the needs of the stakeholders. |
| 66 | + |
| 67 | +=== "Data & Tech Audience" |
| 68 | + |
| 69 | + ## What Is a Concept in the EKG Method? |
| 70 | + |
| 71 | + A **Concept** is the linking pin between local business language |
| 72 | + and formal semantic models (ontologies). |
| 73 | + It enables the EKG to understand business terminology while |
| 74 | + maintaining connections to standardized vocabularies and enabling |
| 75 | + automated reasoning. |
| 76 | + |
| 77 | + ## The Concept Lifecycle |
| 78 | + |
| 79 | + ### Initial Capture |
| 80 | + |
| 81 | + At the initial stages of a use case, the focus should be on |
| 82 | + capturing the language of the users in their domain, which may not |
| 83 | + necessarily involve discussing ontologies. |
| 84 | + The main goal is to gather requirements and understand the |
| 85 | + problem context, as well as the terms and concepts used by the |
| 86 | + users. |
| 87 | + |
| 88 | + !!! tip "Business-first approach" |
| 89 | + |
| 90 | + Start with what the business calls things, not what some |
| 91 | + ontology says they should be called. |
| 92 | + The business owns its use cases and should recognize them |
| 93 | + throughout their lifecycle. |
| 94 | + |
| 95 | + ### Mapping to Ontologies |
| 96 | + |
| 97 | + Later in the use case's lifecycle, once the problem is well |
| 98 | + understood, the relevant ontologies can be mapped to the terms and |
| 99 | + concepts captured earlier. |
| 100 | + This allows for better integration of the use case with the |
| 101 | + overall EKG ecosystem. |
| 102 | + |
| 103 | + ### Refinement and Evolution |
| 104 | + |
| 105 | + As the use case evolves and the understanding of the domain |
| 106 | + becomes clearer, it may be necessary to: |
| 107 | + |
| 108 | + - **Adjust captured concepts** — Better reflect the reality of |
| 109 | + the domain |
| 110 | + - **Map to appropriate ontologies** — Link to more appropriate |
| 111 | + terms in specific ontologies or vocabularies to ensure |
| 112 | + consistency and interoperability across the enterprise |
| 113 | + |
| 114 | + ## Concept Types |
| 115 | + |
| 116 | + Concepts can represent different types of semantic elements: |
| 117 | + |
| 118 | + - **Class Concepts** — Represent types or categories (e.g., |
| 119 | + "Person", "Account", "Transaction") |
| 120 | + - **Property Concepts** — Represent relationships or attributes |
| 121 | + (e.g., "hasOwner", "isLocatedIn", "hasValue") |
| 122 | + - **Shape Concepts** — Represent validation constraints or data |
| 123 | + shapes |
| 124 | + |
| 125 | + ## Concepts as Linking Pins |
| 126 | + |
| 127 | + Concepts are the linking pin in many ways. |
| 128 | + They link "business terms" and "technical terms" as they are |
| 129 | + used in the context of the given use case, by the business and |
| 130 | + by various programs, apps, and systems, in all their variations |
| 131 | + and manifestations. |
| 132 | + |
| 133 | + **Example:** |
| 134 | + The official term for Customer could be "Customer," but it |
| 135 | + appears in different forms: |
| 136 | + - On forms in apps as "Cust." |
| 137 | + - In Python code as `_customer` |
| 138 | + - In SPARQL statements as `?cust` |
| 139 | + - In database schemas as `cust_id` or `customer_id` |
| 140 | + - In API endpoints as `/customers` or `/api/cust` |
| 141 | + - In OWL ontologies as specific axioms (e.g., `owl:Class` or |
| 142 | + `rdfs:subClassOf` relationships) |
| 143 | + - In SHACL shapes as validation constraints (e.g., `sh:property` |
| 144 | + or `sh:minCount`) |
| 145 | + |
| 146 | + All these symbols, terms, axioms, and shapes are manifestations |
| 147 | + of the same concept and linked to it. |
| 148 | + Even the mapping to certain axioms in an OWL ontology or |
| 149 | + certain shapes in a SHACL shape would be seen as "technical |
| 150 | + terms" — just yet some other manifestations of a given concept. |
| 151 | + |
| 152 | + This enables the EKG to understand that these different |
| 153 | + representations all refer to the same business concept, enabling |
| 154 | + semantic integration across diverse systems and technologies. |
| 155 | + |
| 156 | + ## Relationship to Ontologies |
| 157 | + |
| 158 | + Concepts serve as a bridge between: |
| 159 | + |
| 160 | + - **Local business terminology** — The language used within a |
| 161 | + specific use case or domain |
| 162 | + - **Technical manifestations** — How concepts appear in code, |
| 163 | + databases, APIs, forms, and other systems |
| 164 | + - **Standard ontologies** — Formal semantic models that enable |
| 165 | + interoperability and reasoning |
| 166 | + |
| 167 | + This multi-faceted nature allows the EKG to: |
| 168 | + - Address "the business" with their language |
| 169 | + - Integrate with technical systems using their terminology |
| 170 | + - Keep backend EKG models generic and linked to appropriate |
| 171 | + ontologies |
| 172 | + - Model semantic conundrums (e.g., different terms for the same |
| 173 | + concept, same term for different concepts) |
| 174 | + |
| 175 | + ## Reuse and Inheritance |
| 176 | + |
| 177 | + Concepts follow the same reuse patterns as other elements in the |
| 178 | + EKG Method: |
| 179 | + |
| 180 | + - **Local definition** — Each use case can define concepts |
| 181 | + specific to its domain |
| 182 | + - **Inheritance** — Lower-level use cases inherit concepts from |
| 183 | + parent use cases in the Use Case Tree |
| 184 | + - **Borrowing** — Use cases can reference concepts from related |
| 185 | + use cases |
| 186 | + - **Consistency** — Common concepts are defined once and reused |
| 187 | + |
| 188 | + This ensures that concepts are not duplicated unnecessarily and |
| 189 | + that the EKG maintains semantic consistency across the enterprise. |
| 190 | + |
| 191 | + ## Relationship to Other Concepts |
| 192 | + |
| 193 | + Concepts are fundamental building blocks that relate to: |
| 194 | + |
| 195 | + - **[Use Cases](use-case.md)** — Each use case has a vocabulary |
| 196 | + of concepts |
| 197 | + - **[Personas](persona.md)** — Personas are Concepts, enabling |
| 198 | + semantic definition and reasoning |
| 199 | + - **[Stories](story.md)** — Stories reference domain concepts |
| 200 | + that need to be understood and modeled |
| 201 | + - **[Ontologies](ontology.md)** — Concepts link to ontology |
| 202 | + classes, properties, and shapes |
27 | 203 |
|
0 commit comments