search

swiss-art-research-net / reference-data-models

Reference Data Models discussions and integrations
8 stars 1 forks source link

Diversity in patterns #18

Open ncarboni opened 4 years ago

ncarboni commented 4 years ago

When patterns are different?

  1. E1 --> P1 --> E41

  2. E1 --> P1 --> E41 --> P2 --> E55(x)

  3. E1 --> P1 --> E41 --> P2 --> E55(y)

  4. E1 --> P1 --> E41 --> rdf:value --> rdf:Literal

  5. E1 --> P1 --> E41 --> rdfs:label --> rdf:Literal

  6. E1 --> P1 --> E41 --> Px_has_symbolic_content --> rdf:Literal

In my perspective, these are all different patterns. Specifically,

  1. E1 --> P1 --> E41 --> P2 --> E55(x)
  2. E1 --> P1 --> E41 --> P2 --> E55(y)

Example 2 and 3 do have similar structure, but the E55 change the meaning of the node (E41) they are attached to, therefore making them semantically different.

  1. E1 --> P1 --> E41 --> rdf:value --> rdf:Literal
  2. E1 --> P1 --> E41 --> rdfs:label --> rdf:Literal
  3. E1 --> P1 --> E41 --> Px_has_symbolic_content --> rdf:Literal

Examples 4, 5, 6 do have express the same semantic content but it is expressed differently, so they are syntactically diverse.

ncarboni commented 4 years ago

Agreed that example 2-3 represent the diverse instantiation of the same meta pattern. The instantiation are important and are the one showed to the users (e.g. preferred and alternate appellation), and should be documented at a field level with separate it.

Examples 4, 5, 6 are different