Datamap specification

[HLWeil]

This PR adds the specification of the Datamap (Former Working name Dataset) to the ARC-specification. The main idea behind the Datamap is to give additional context and explanations about the data entities defined in an assay or a study. To keep this streamlined, it also makes use of referencing the data entities of interest using the Data fragment selector principle. The motivation behind this additional context is to allow for a more static description of the data entities besides referencing them out of the processes defined in annotationTables. This will add a lot more flexibility to power users on the one hand, and an easier point of access for new users on the other hand.

I propose the Datamap as an extension only in the ISA-XLSX specification here, as the extension of the ISA abstract model was not accepted: ISA-tools/isa-specs#15

The extension adds three parts:

• Definition of the Datamap table sheet format containing details about the different columns (ISA-XLSX.md)
• Defintion of the Datamap file format (ISA-XLSX.md)
• Reference of the Datamap file in the Study and Assay section of the (ARC-specification.md)

I dropped the name Dataset, as this is already being occupied by the Dataset type in Schema.org, which is a first-class citizen in RO-Crate. The alternative name Datamap is debatable. I chose it over DataDictionary because it's much shorter but covers roughly the same semantic principle. Also, DataContext might be a fitting alternative. Furthermore, we might change typing to DataMap.

Input would be very welcome, @kappe-c @floWetzels @ZimmerD @kMutagene @Freymaurer @chgarth @muehlhaus @gdoniparthi

closes #80
closes #82