FAQ
All you need to know to use muzAI
This platform has been developed and tested with theChromebrowser. We strongly recommend this browser when using our platform.
You have to upload ZIPPED datasets that can contain 1 or more files.
Datasets must be under the following formats: CSV, XML (currently working on JSON format)
If you import XML datasets compliant with the EDM model, an automated mapping will be applied (currently working an automated mapping for LIDO).
If you import a CSV dataset, you can use a specific naming convention for its header in order to benefit from multilingual features and to allow more than one image to be associated with each document/line. Please use thisCSV templateto check how to implement this into your data.
If you import a XML dataset, you can use our mapping tool to include language attributes in the data conversion.
If you import a XML dataset, you can refine how its data is imported by manually editing the list of XPATH gathered in its dedicated mapping.
The system will consider only items carrying an identifier, an image and either a title or a description. An item missing one of these requirements will be discarded (you can download separately these discarded files for your information).
The processing of your datasets is done in 3 steps
- Itemizing and fetching the original images : based on its mapping, a dataset is first itemized and then, all its related images are fetched from their provided URLs. After this step is finished, you might have to re-trigger it again if some content has not been retrieved (this can be related to various reasons, such as servers downtime or invalid URLs that might arise).
- Manifests generation : Then, the dataset items are transformed into IIIF manifests, while concepts and landmarks are automatically extracted from their images. The export to EDM dataset is also happening during this step.
- Preview and publish : Finally, you can preview the results in a restricted environment before publishing into the public space dedicated to your institution.
Your collections are managed under a main organisation/company that you need to create in a first stage. This entity will be used as your public representation, and can be officially authentified if you wish.
Once published its first collection, your institution will be given a public landing page, from which all the related collections can be accessed.
Your collections, once published, will be available in the Overview/Explore sections of the muzAI platform with all the other collections published by our other users.
If you edit your organisation or collections details, you will have to unpublish and re-publish the latter to update their related objects metadata.
Once your datasets have been fully processed and validated, you will have access to the automatically exported EDM dataset as a separate download.
In a later stage, we will propose an API to retrieve programmatically the IIIF manifests too.