4.4 Metadaten

Im Punkt 9 der OGD-Prinzipien (s. Dokumentation (Dauerhaftigkeit)) ist geregelt, dass veröffentlichte Daten mit umfangreichen Metadaten zu dokumentieren sind vgl. COGDÖ 1 2012: 8. Die Cooperation OGD Österreich schuf die Arbeitsgruppe Metadaten, um ein einheitliches Vorgehen in Österreich gewährleisten zu können. Diese Arbeitsgruppe brachte am 12.03.2012 ein entsprechendes Dokument heraus COGDÖ 2 2012.

Die Metadatenstruktur in Österreich basiert auf zwei Bereichen. Das ist einerseits der Metadatenkern mit 12 Pflichtfeldern. Ergänzend dazu sind 18 alternative Attribute vorgesehen, wobei die Gesamtstruktur erweiterbar ist. Zusätzlich wurde die Metadatenstruktur auch in die englische Sprache übersetzt, um international kompatibel zu sein. Wenn es sich um INSPIRE-Datensätze handelt, werden auch die INSPIRE-Metadaten ausgelesen, umgekehrt sind nicht automatisch OGD-Datensätze in INSPIRE gelistet vgl. COGDÖ 2 2012: 6ff.

Um Datensätze nicht nur maschinenlesbar (Punkt 5 der OGD-Prinzipien), sondern „maschinenverstehbar“ zu gestalten, sind diese mit semantischer Information anzureichern um weitere Vorteile zu bieten: „When PSI is converted into LOD, it creates interesting possibilities for analyzing policies of public bodies. These analyses are important because they create transparency of policies and the possibility to validate the open data and to examine the open data for hitherto undiscovered relationships. […] But if one wants to use LOD to analyze policies, the understanding, use, and management of these data should be facilitated. For instance, one needs to know how the data were created and how they should be interpreted. This type of information is denoted as metadata. […] In the ideal situation for LOD, different types of metadata are provided: discovery (flat) metadata (which are descriptive and navigational), contextual metadata (which are descriptive, restrictive and navigational) and detailed metadata (which cover schema metadata plus additional metadata to assure quality).Zuiderwijk 2012: 282f.

Welche Aufgaben Metadaten überhaupt haben und wie diese im Lichte von Linked Open Data zu bewerten sind, haben Zuiderwijk et al. untersucht (Tabelle 6).

Benefits of metadata according to literature overview Challenges derived from the use cases
Create order within datasets Provide structured and sufficient metadata.
Improve storing and preservation of LOD Provide metadata so that explicit knowledge can be derived from the LOD on the long term.
Improve easily finding LOD Provide search possibilities based on metadata and a persistent identifier of the LOD so that certain data/researchers can easily be found.
Improve the accessibility of LOD Provide metadata that make it easy to locate and retrieve the data (e.g. search possibilities) and make it possible to request special permission to be able to access the data.
Assess and rank the quality of LOD Provide metadata that can be used to assess and rank the quality of LOD.
Improve easily analyzing, comparing, reproducing and finding inconsistencies in LOD Provide metadata about how the research was conducted: e.g. how were the data analyzed? Which choices were made during the research?
Improve chances of a correct interpretation of LOD Provide metadata that explain the variables that were used for the research, the choices that were made during the research, show relations with other data/websites.
Improve possibilities to find patterns in LOD Provide metadata in different formats (e.g. XML, CSV, etc.), so that their interoperation allows for patterns to be detected in different LOD datasets, even if the syntax and semantics are different.
Improve visualizing LOD Provide extensive metadata about the quality (e.g. accuracy) of the data that make it easier to visualize data (e.g. for data mapping). Make linking data easier Provide metadata that describe the variables extensively and that provide knowledge about which variables could possibly be linked. Provide metadata about the data providers and make it possible to approach them for questions about the LOD (e.g. by E- mail).
Avoid unnecessary duplication of LOD Provide access to data and metadata and show similarities with other data.

Verwendete Literatur

Tags

Interessante Dokumente