...
While performing this step, keep your
...
For easier understanding, we will follow the example of the below set of metadata elements. The steps we are following in our example can be seen with the red arrows in the flowchart above.
Metadata Field | Value |
|---|---|
Dataset Name | Health Data |
Date of Upload | 01/02/2023 |
Keywords | BP, HR, Conditions |
Creator | Dr. Smith |
Description | Patient health data including BP and HR |
Format | CSV |
Source | Hospital A |
Rights | Open |
| Panel | ||
|---|---|---|
| ||
Example - step part 0: In this example, we are working with existing metadata. According to the flowchart, we should start with Check for an existing standard/codebook. |
Step: Compile information
Compile all the information of (meta)data elements, (meta)data values, and (meta)data structure. Examine the (meta)data in the way it is currently stored, including its format (e.g. JSON, CSV) and how the information is organized within it. This step helps to identify inconsistencies, ambiguities, and errors in the data.
...
Which variables are present in the (meta)data?
For example, check the eCRFs used to collect the data.
What are the value ranges for each variable, as in determine the type of values each variable can have?
For example Dataset Name has a range of Text (e.g., "Health Data").
What are the allowed value ranges for data elements.
For example, for the variable ‘sex', are only ‘male' and 'female’ present, or also ’intersex'?
Are there any existing relationships between the variables in your (meta)data and do they make sense?
For example, the “sex” variable is an attribute of “patient”, which may imply that the semantics of this “sex” variable is “sex of patient”.
Example - step part 2: In our example, we are working on FAIRifying an already existing metadata of a dataset. Let’s compile and examine the information we have.
...
You should end up with a compiled table of information, like you can see in the table below step a) and continue to Step Check for an existing standard/codebook(for new (meta)data).
Step: Check for an existing standard/codebook
a) For existing (meta)data
...
If the (meta)data is not coming with a codebook or standard, proceed to Compile information.
Example - step part 1: In our example, we are working on FAIRifying existing metadata. We have checked our metadata, but it does not come with a metadata standard, so we’ll have to compile all information of our existing metadata first.
...
| Info |
|---|
Health-RI, together with domain representatives, will be aiming to develop domain-specific national (meta)data standards in the future. |
Example - step part 3: After having compiled all elements in our metadata, let’s say we explored the relevant repositories and, for the purpose of following the example through all of the steps, couldn’t find one that would be applicable to our example metadata. According to the flowchart, we should now move to the Check and improve (meta)data semanticsstep.
Step: Check and improve (meta)data semantics
In order to end up with clearly defined (meta)data elements, we should check the semantics to identify things that are unclear or ambiguous. Next, we can improve these ambiguities. For this, first check:
...
This recipe in the FAIR cookbook gives some additional guidance on specifying the semantics of elements of your data.
Example - step part 4: In the below spreadsheet we can see what the issues are with our current metadata and applied improvements in order to make the meaning of them clearer.
For example, we split the element Creator into Creator Name and Creator Identifier, to improve the identification of the Creator. We also defined the format in which the Date of Upload should be stated to prevent confusion with dates/months.
Metadata Field | Original Value | Issue | New variable description | New Value Range | New Value |
|---|---|---|---|---|---|
Dataset Name | Health Data | Generic and not descriptive. | The name of the dataset. | Text | Patient Health Records 2023 |
Date of Upload | 01/02/2023 | Ambiguous format | Date when the dataset was uploaded, in ISO 8601 format (YYYY-MM-DD). | Date in ISO 8601 format (YYYY-MM-DD) | 2023-01-02 |
Keywords | BP, HR, Conditions | Abbreviations used without context. | Keywords describing the main topics covered by the dataset. | Text | Blood Pressure, Heart Rate, Hypertension |
Creator Name | Dr. Smith | Generic name without additional identifying information. | Full name of the dataset creator. | Text | Dr. John Smith |
Creator Identifier | None | Missing in the original example metadata, we need the identifier to make sure we refer to the correct agent. | ORICD of the creator. | ORCID identifier | ORCID: 0001-0002-3456-7890 |
Creator Institute | None | Missing in the original example metadata, we added it to make the creator more identifiable. | Affiliation of the dataset creator. | Text | Hospital A |
Description | Patient health data including BP and HR | Lacks detail. | Extended description providing context and details about the dataset. | Text | Detailed patient health records including measurements of blood pressure (BP) and heart rate (HR), along with diagnosed medical conditions and prescribed medications. |
Format | CSV | None | Data format of the dataset. | Text | CSV |
Source Name | Hospital A | Lacks detail, too generic. | Specific department and institution where the data was sourced. | Text | Hospital A, Department of Cardiology |
Source Identifier | None | Missing in the example to unambiguously identify the institute. | ROR identifier that is specific to the (research) institute. | ROR identifier | |
Rights | Open | Too broad. | Licensing terms specifying the rights for data usage. | URL to CC License |
Step: Go to the next Metroline Step
Congratulations, you have now successfully analyzed your (meta)data semantics. You should now have a set of data elements (variables) with clear and unambiguous semantics - a codebook. For metadata, you should be left with a set of clearly defined metadata elements.
You can now continue to the next step: Metroline Step: Create or reuse a semantic (meta)data model, which uses work done in this step as a basis to find correct ontologies that can be coupled to your data elements.
Example - step part 5: Great job! After checking and improving our metadata semantics, we can now move to the next Metroline step.
...