Component Metadata Infrastructure
Menzo Windhouwer, Twan Goosen
Abstract
At the start of CLARIN, metadata for language resources faced various problems, for example, different communities using different terminology, which made interoperability difficult. The Component Metadata (CMD) Infrastructure (CMDI) was developed as a solution and is based on specifying reusable components, each of which contains other components and elements. Components are assembled into profiles, the schema for the metadata description of a specific type of language resources. The CMD Infrastructure consists of many interacting parts, including a Component Registry, several semantic registries, a metadata harvester, and a central catalogue (the VLO). It is supported by repository systems and metadata editors developed and maintained by various stakeholders across the CLARIN network. The CMDI landscape has expanded throughout the years, and has remained sustainable by adapting itself, as it will continue to do in the future.