coli-conc Logo

A novel approach to terminology mappings

coli-conc Technical Report 3

2016-04-01 CC-BY-SA

Jakob Voß <> (Verbundzentrale des GBV (VZG))


The need of mappings between terminologies, vocabularies, ontologies, and other kinds of knowledge organization systems has been articulated since introduction of such systems at large scale.[1] For this reason project coli-conc aims at developing an infrastructure to facilitate management and exchange of (cross-)concordances between terminologies. This report outlines a novel architecure of terminology mappings to ultimately connect all knowledge organization systems.


As soon as concordances cross more than two vocabularies, multiple architectures can be applied [@ISO25964-2]. For small numbers, a many-to-many architecture is feasable, but for more terminologies a hub architecture becomes more appropriate [@Binding2015; @Soergel2011]. This requires people to agree on a common hub, but people tend to not agree especially on issues of knowledge organization. Networks of cross-concordances have been applied with success [@Mayr2008] but web-scale applications of distributed mappings require a new architecure.

Many-to-many architectureHub architecture

The noodle architecture

We propose the "noodle achitecture" as opposed to the many-to-many architecture and the hub architecture for managing a large number of concordances. In our model each knowledge organization system is once mapped to one other system. This second system is not a central hub but just another terminoloy that happens to be the last one not connected with two other terminologies yet. The resulting data structure has the shape of a very long noodle.

Noodle architecture

This architecture scales as well as the hub architecture but no central terminology is required. We expect this to result in much less debate among information architects because all terminologies are treated equally. To ensure that only one terminology is added at a time at the end of the noodle, the whole data structure is managed as blockchain: this kind of distributed database known from Bitcoin can be applied to other domains as well [@Nakamoto2008].


This report introduces a novel approach to manage a very large[2] number of mappings between any kind of knowledge organization systems. The so called noodle architecture combines the benefit of a hub architecture with less struggle about which system to use as central hub.


  1. For instance at the 1910 conference of the International Federation for Information and Documentation (FID). ↩︎

  2. In notions of big data as explained by @Shaw2015. ↩︎


Introduces tagging mismatch reasons and non-indexing concepts


Improves display of concept information, and many other small changes


Adds support for new filters, suggestions via catalog co-occurrences, and additional vocabulary APIs


Sacherschließungsdaten des K10plus-Katalog als Open Data veröffentlicht


The software library cocoda-sdk can be used to access terminologies from various sources


Informationen zum Einbinden eigener Vokabulare in Cocoda anhand verschiedener Beispiele.