Migrating or exporting data from a VRE
This page describes the general process of moving data out of SharePoint and into a data model and data format that can be used in different contexts. In general, these converted data can be readily stored in a data archive.
If you have a SharePoint VRE and would like to export your data, please contact the Centre for Digital Scholarship for information.
Data modelling¶
Unless you only used the VRE for sharing files that are read and edited outside the VRE, you probably have information that is managed directly in SharePoint. Such data may be stored in lists. The data model explains what the information is about.
Some projects already started with a highly structured data model, other projects need more help to reshape their data into models that are more interoperable with existing datasets.
In many cases, we will meet several times to discuss what your data describe and how data points are interrelated.
Find models to map to¶
For your data to be understandable to other researchers (or anyone else, really), it helps to "speak the same language", i.e. use common data models. Many people have already created data models to describe for example books, people, architecture and art, genes, et cetera. However, not everyone describes every item in the same way, so you may need to extend or adapt an existing data model.
There are too many possible models to list here, so we only link to some databases that provide overviews of existing models:
- Linked Open Vocabularies (LOV) indexes ontologies and RDF Schemas
- FAIRsharing collects standards, databases and policies
Iteratively create mapping(s) and test their correctness¶
Use SharePoint's legacy XML export, REST API with Atom XML or download options in the web interface to get the data.
Transform the data to the target model using, for example, XSLT or RML. Do not try to create a complete mapping in one go, but start simple and improve one step at a time.
Perform the migration¶
When you are happy with the results of your mapping and transformation testing, do it once more for all data and add metadata about the dataset, including its provenance. This is the data that you will publish, archive or maybe continue working on using other tools.
Archive the data¶
There are many services for archiving datasets, usually known as data archives. See the Research Data Services catalogue for information on data archives that we know of (though this catalogue is incomplete and may be out of date).