Store & Manage Data
To ensure that research data remains findable and usable in the long term, it should be well structured, described with meaningful metadata, and stored in an appropriate infrastructure.
Metadata
Metadata is data about data. It plays an important role in making information findable, understandable, and reusable. Metadata consists of two components:
- Schema = Metadata Key. Schemas define what information is recorded.
- Value = Metadata Value. Vocabularies define how the information for each metadata key should be specified.
Example:
| Metadata Key | Metadata Value |
|---|---|
| Author | Max Mustermann |
| Date | 01/01/2001 |
| File format | PDF/A |
| Language | en |
| License | CC-BY 4.0 |
Metadata should be human- and machine-readable so that it can be easily understood, efficiently processed, and linked. Ideally, metadata is captured when the data is created and updated regularly thereafter. There are several ways to store metadata:
- directly with the data
- in a documentation tool
- in a separate file stored alongside the data it describes
Metadata can be documented in a README file (.txt, .md), in tabular form (.csv), in structured formats (.json, .xml), or in the input fields of a research data platform such as TUM DataTagger.
Metadata standards help ensure that data is described and exchanged consistently. They can be generic, such as DataCite, or discipline-specific. The level of detail required depends, among other things, on the target audience, collaboration with project partners, and where the data will be published. Metadata registries can help you find a suitable metadata standard for documenting your data. Examples include FAIRsharing.org and the RDA Metadata Standards Catalog.
File Organization
Structured file organization can be achieved through both metadata and a hierarchical folder structure. Depending on your needs, the folder structure can be organized by topic, workflow, or file type.
Files within a folder can be sorted, for example:
- alphabetically or by topic: 001_Topic_X, 002_Topic_Y, …
- by date: YYYYMMDD_Topic_XY
File names should be meaningful, consistent, and free of special characters and spaces. Especially when working in a team, it is recommended to document the agreed file naming convention in a README file.
In systems without a hierarchical folder structure, such as TUM DataTagger, assigning metadata is essential for finding specific files. Information from an existing folder structure can be mapped to the tool’s metadata fields.
Infrastructure
Using digital tools and suitable, well-planned infrastructure can help you manage your data. Follow this decision tree to find a suitable infrastructure solution. You can choose from storage solutions, infrastructure for sharing or collaboratively editing files, as well as tools for documenting data or research processes. For detailed information on the infrastructure services available at TUM, visit the Tools & Storage Solutions page.
Decision Tree for Finding a Suitable Infrastructure Solution
Get in touch!
For questions on data storage and management, please contact us at researchdata@tum.de!
Further information: Submit application & planning, Analyse data, Archive data, Publish & share data, Reuse data