Hi Joao

So I asked one of our MOOC contributors Katarzyna Biernacka to answer this, as she is a research data expert: 

To achieve this, metadata and data should be well-described. This means: write a documentation and use metadata standards, controlled vocabularies, ontologies, thesauri and authority files.

To find metadata standards you can go to http://www.dcc.ac.uk/resources/metadata-standards or http://rd-alliance.github.io/metadata-directory/subjects. Usually, as a researcher you don't really know what standard is used in a repository, you just fill in the form. But if you are interested you should have a look - it helps as well to choose the right vocabulary. To find the right vocabulary you can go to bartoc.org or to taxonomywarehouse.com and search for it depending on your subject. Sometimes it's hard to find something; be patient.

Another very important thing are clear use rights. Provide licenses to your data to make it re-useble. You are the creator of the data and you can let other use your data. Try to use CC0, CC BY or CC BY-SA regarding to Open Science. Remember though, not to all data licenses can be applied - it need to be data that has some intellectual creation in it.

(See attached documents for more guidance, one is in German but should still be useful).