Open Data Protocols, DCIP [A Topic Map Song]

Open Data Protocols, DCIP

From the post:

Have you ever heard about Data Catalog Interoperability Protocol (DCIP)? DCIP is a specification designed to facilitate interoperability between data catalogs published on the Web by defining:

  • a JSON and RDF representation for key data catalog entities such asDataset (DatasetRecord) and Resource (Distribution)based on the DCAT vocabulary
  • a read only REST based protocol for achieving basic catalog interoperability

Data Catalog Interoperability Protocol (DCIP) v0.2 discusses each of the above and provides examples. The approach is designed to be a pragmatic and easily implementable. It merges existing work on DCAT with the real-life experiences of “harvesting” in various projects.

To know more about DCIP, you can visit the  Open Data Protocols  website, which aims to make easier to develop tools and services for working with data, and, to ensure greater interoperability between new and existing tools and services.

The news of new formats, protocols and the like are music to topic map ears.

The refrain is: “cha-ching, cha-ching, cha-ching!”


Only partially in jest.

Every time a new format (read set of subjects) is developed for the encoding of data (another set of subjects), it is be definition different from all that came before.

With good reason. Every sentient being on the planet will be flocking to format/protocol X for all their data.

Well, except that flocking is more like a trickle for most new formats. Particularly when compared to the historical record of formats.

In theory topic maps are an exception to that rule, except that when you map specific instances of other data formats, you have committed yourself to a particular set of mappings.

Still, that’s better than rip-and-replace or ETL processing of data. It maintains backwards compatibility with existing systems while anticipating future systems.

2 Responses to “Open Data Protocols, DCIP [A Topic Map Song]”

  1. says:

    Ye gods. I take a set of data about some entity (myself, Oslo, Munch’s The Scream, a nearby river, or whatever), and then, because I publish it, it becomes an instance of DatasetRecord? How that’s going to help anyone find anything is beyond me. Seems pretty much equivalent to pronouncing the thing a subject in Topic Maps terms, which is equivalent to saying nothing at all.

  2. Patrick Durusau says:

    Tis true, tis true.

    What I hope for is the use of topic maps to annotate useless instances in hopes of them becoming more useful. Not hard considering their original utility. 😉