This is a tool for the inspection of the metadata types in a collection, as well as for the evaluation of metadata quality of those fields. At this moment (Version 3), the Inspect tool offers the option to evaluate metadata completeness, which is calculated based on the number of records for which a given metadata field has been filled in. This completeness can be evaluated per field (all years), or along time.
The “collection inspection” functionalities support the scholarly tasks of data criticism by facilitating the close inspection of the metadata fields that constitute each collection.
The purpose of this tool is to provide an overview of how a collection/dataset is constituted and also to allow a closer inspection of their metadata (e.g., detect incomplete data, or observe value distributions along date fields).
It is important to have clear that the collection inspector does not give search results. Future developments include the integration of other metrics (besides completeness) for the evaluation of metadata quality, and the possibility to visualize metadata completeness together with search results.
If you are interested in the completeness of the ASR (automatic speech recognition transcripts) of the Netherlands Institute for Sound and Vision’s audio-visual collection, visit this page .
Because the Media Suite has been built following the principles of metadata transparency (which means that we only a minimum intervention in normalization and cleaning is done to the metadata provided by the institutions), users may face difficulties in understanding what a specific metada field means, or how to use it.
In Version 3, we provide an example of a good practice in documenting the metadata schemas for the users. This is done via “metadata dictionaries,” which provide definitions of the metadata field (regardless of their labels) when they are used in the Media Suite.
We have integrated this dictionary for one collection (The Netherlands Institute for Sound and Vision). The institution was responsible for creating a file with the technical label of the field plus a user-friendly label and a user-oriented definition of the field. These data was integrated into the Media Suite (you can observe it, for instance, when you click in the drop-down menu in the Inspector tool.
The CLARIAH WP5 team encourages other providing institutions to create and publish these metadata dictionaries as part of their metadata schemas and documentation.