README.md

Simplified Semantic Data Modeling (S2DM)

S2DM is an approach for modeling data of multiple domains. It is simple in the sense that any Subject Matter Expert (SME) could contribute to a controlled vocabulary with minimal data modeling expertise. Likewise, it is semantic in the sense that it specifies meaningful data structures, their cross-domain relationships, and arbitrary classification schemes.

Note

Bear in mind the word Simplified in the name. This approach aims to foster the adoption of (some) good data modeling practices. It does not intent to re-invent, nor to replace long-standing standards, such as those of the Semantic Web. Hence, this approach does not incorporate advanced reasoning capabilities or the use of comprehensive ontologies typically associated with traditional semantic data modeling.

S2DM adopts data modeling best practices and reuses the following elements:

• GraphQL Schema Definition Language (SDL). It provides a clear, human-readable syntax for defining data structures and relationships, making it easy for SMEs to understand and use without requiring deep technical expertise.
• Simple Knowledge Organization System (SKOS). It offers a straightforward framework for creating and managing hierarchical classifications and relationships between concepts, facilitating the organization and retrieval of knowledge in a way that is both intuitive and semantically rich.

To learn more about the background that has led to S2DM, as well as its design principles, please read the S2DM Approach Primer.

Basic principle

The idea is that multiple systems in the physical layer (e.g., databases, streaming platforms, applications, etc.) can share the same concepts. However, instead of arbitrarily modeling domains in the physical layer, the purpose is to have a unique way for specifying the concepts of common interest and its organizing principles in such a way that they are reused. This principle is a core part of a data-centric architecture, reducing undesired duplications and software waste when it is systematically applied.

Loading

graph LR
  subgraph ConceptualLayer
    Person
    Vehicle
    ParkingLot
  end
  subgraph PhysicalLayer
    Database
    StreamingPlatform
    Application
    Other
  end
  ConceptualLayer --> Database
  ConceptualLayer --> StreamingPlatform
  ConceptualLayer --> Application
  ConceptualLayer --> Other

In this sense, S2DM is an approach to specify those concepts of interest systematically. A more generic (and elaborated) diagram looks like the following:

Loading

graph LR
  subgraph Conceptual layer
    spec_file_1.graphql
    spec_file_2.graphql
    spec_file_N.graphql
    subgraph S2DM Tools
      Composer
      Exporter
      Other
    end
  end
  subgraph Physical layer
    App_SHACL
    App_YAML
    App_JSON
    App_Other
  end
  spec_file_1.graphql --GraphQL schema 1--> Composer
  spec_file_2.graphql --GraphQL schema 2--> Composer
  spec_file_N.graphql --GraphQL schema N--> Composer
  Composer --Merged GraphQL schema--> Exporter
  Composer --Merged GraphQL schema--> Other
  Exporter --VSPEC--> App_YAML
  Exporter --SHACL--> App_SHACL
  Exporter --JSON schema--> App_JSON
  Exporter --Other?--> App_Other

Getting started

The S2DM consists of the following artifacts:

• S2DM data modeling guideline - It explains how to formalize the data of a domain with the S2DM approach. In other words, how to create the specification files that will constitute the core of the conceptual layer. Click HERE to start modeling a domain with S2DM approach.
• S2DM Tools - Code that support the proper usage of the S2DM guideline. It helps with the modeling language validation, identifiers, search functions, exporters, etc. Click HERE to start using the S2DM tools.

Contributing

Contributions are welcome. Click HERE to contribute (to the S2DM approach itself).