Data sets are represented as Objects on SMIR. An object can consist of multiple files (e.g. DICOM series) or a single data file (e.g. statistical model). Objects can be related to each other. Related objects are visible in the Related data section of each object. After upload, the objects are only visible to the owner. These objects are called unpublished data. After curation by the owner (filling meta-data, reviewing the data, setting permissions), the data has to be pulished to make them available to other users. Each object has a unique SMIR-ID and is renamed according to the SMIR naming template. The original file name is only visible to the user who owns the data.
Example for ID 101597
There are different object types to represent the type of the data set. Example object types are:
- RAW image like CT scan, MR images, etc.
- Segementation of images
- Surface models
- Genetic data
For a complete list check the Supported Data Section.
SMIR uses a template to replace filenames after upload. This helps to identify what type of data the object represents by just looking at the name. It also removes identification information stored in the original (local) filename. The original filename is stored but only displayed to the owner of the object.
- Namespace: SMIR
- Anatomy: Right scapula
- Age: 56
- Gender: male
- Modality: Other Type (OT)
- ID: 167141
- Extension: STereoLithography (stl)
General meta data
Type specific meta data
Export meta data
Meta data of an object can be exported on the detail page. Available formats:
- SMIR API (json)
- Dublin core (xml)
SMIR signed an agreement with a DOI Registration Authority (ETH Zurich) to enable attaching a Digital Object Identifier (DOI) to any dataset published on the database. Building and curating data collections is difficult and time consuming. This permanent identifier will increase the visibility of the work by making reference to your datasets easy. Furthermore, the DOI published on SMIR are automatically registered on DataCite – a non-profit website, where these datasets are accessible for search based on their meta-information.
! A DOI identifies an object permanently, even if the object changes its location, its owner or other characteristics. The DOI does this through redirection, using the Handle system.
A Create DOI button is visible for all published objects and can be used to request a DOI.
A confirmation is needed in order to initate the DOI creation process.
After you confirm the DOI creation:
- SMIR creates a valid DOI. Visible in the File Information Section.
- SMIR exposes the DOI and meta data to the Registration Agency (ETH Zurich).
- The Registration Agency collects the data and registers the DOI with DataCite.
- After 48 hours the DOI is active.
The access to each object is defined based on fine-grained permissions.
- Visit (V) allows the associated user/group to see the object (e.g. in a folder).
- Read (R) allows the associated user/group to read the metadata of the object.
- Download (D) allows the associated user/group to download the object.
- Edit (E) allows the associated user/group to edit the object.
- Manage (M) allows the associated user/group to manage the object (same permissions as the owner).
A permission set is a combination of multiple object permissions.
- Protect allows the associated user/group to see the object and read the metadata.
- Default allows the associated user/group to download the object including the permissions of Protect.
- Collaborate allows the associated user/group to edit the object including the permissions of Default.
- Private means the associated user/group is explicitely not allowed to see the object at all.
During the publishing procedure, the following default permissions are applied to each object. At any point it is possible to change or remove the permissions entirely as an owner of the object.
Based on the default permissions all visitors (registered or not) know about the existence of the object which means they are allowed to read the metadata. However, only registered users are allowed to download the object.
Ech object has exactly one owner. But the owner can change (transfer) the ownership of the object(s).
- Owner change does not affect existing permissions!
- You will lose control over the selected objects!
Search the new Owner and confirm the change in the modal to finish the process
- Select multiple objects using the checkboxes.
- Select Owner in the Manage selected drop-down menu.
- Search the new Owner and confirm the change in the modal to finish the process.
The license specifies the rights the user has after download whereas object permissions define the access to the file on SMIR.
A license is set for each object. Copyright is set as default. Please consider changing it when you prepare the object for publishing. The license file will be included in the download. A list of open source license are available. If you need a custom license please use the ticket system to send such a request.
Change license of a single object
The user can change the license in a object list view or on the object's detail page.
- Navigate to the object
- Click on the menu icon (list view)
- Select Edit license
- Select the license from the drop-down menu in the modal
- Set license to change or cancel to abort
Batch license change
- Select the object and click on
Manage selectedand then
- Select the license from the drop-down menu in the modal
- Set license to change/save or cancel to abort
SMIR generates for Images and model a set of up to 10 preview images to give a glance at the data set. Preview images are not part of the actual data you download.
The actual data is not displayed/accessable on SMIR. To use the data set, download the object. However, the list of included files in an object is displayed.
Download are wrapped in a directory contains the following files and downloaded as compressed archive (zip)
- Data files
- Meta data file: .json
- Readme README.md
- SMIR citation as Bibtex: smir.bib
- License: License_*.txt