Add new class dcat:DatasetSeries to DCAT-AP CH

[mispichtig]

Class: dcat:DatasetSeries
Property: at least mandatory and recommended properties
Implementation:
Conformance: dcat:DataSeries is an optional class of DCAT-AP 3.0.0
Proposal: Introducing the class dcat:DataSeries with at least mandatory and recommended properties

Introduction of dcat:DatsetSeries in DCAT-AP v 3.0.0: https://semiceu.github.io/DCAT-AP/releases/3.0.0/#DatasetSeries

Important note on alignment with DCAT 3: https://semiceu.github.io/DCAT-AP/releases/3.0.0/#notes-on-alignment-with-dcat-3

The introduction of Dataset Series in DCAT is an additive operation from semantical, data modeling perspective. Therefore the impact on existing DCAT-AP 2.x exchanges is limited. Nevertheless, publishers may face substantial impact on the existing catalogues as the new terminology for Dataset Series may require to revisit the publisher's metadata guidelines. For instance, if the publisher had chosen to document a Dataset Series as a Dataset with multiple Distributions, then the alignment with Dataset Series as expressed in this specification will require a substantial effort. This impact goes beyond the specification and dependent on the usage of the shared information. In the context of DCAT-AP and in line with the conformance expectations, the term Dataset Series will be solely used for resources that are explicitly identified by the class Dataset Series. The situation, as mentioned, where a Dataset has multiple Distributions, will be considered as a Dataset with multiple Distributions and not as a Dataset Series. This way, receivers of DCAT-AP metadata can rely on the classes to process the metadata.

Discussion about Using a Dataset Series as a Dataset with multiple Distribution (Antipattern)
dcat:Distributions Usage note should not encourage the usage of Distributions as Dataseries · Issue #112 · opendata-swiss/dcat_ap_ch (github.com)
Diskussion: Model Series of Data as Distributions of a single Dataset · Issue #1429 · w3c/dxwg (github.com)

=> Important note: we should avoid doing our own thing. But how do we handle datasets that use distributions as series?