In the “Collection stats” it might be more informative to have the description of the collection, even the link to the registry, instead of the text “In the tabs below it is possible to inspect the …”. A shorter version of this text can be shown on mouseover over the chart icon.
To use the same terminology in the Collection inspector guidelines, I understand this refers to the static text in the “collection stats” that opens after clicking on the “metadata inspector” icon. Since this suggestion requires some user evaluation to check how to integrate it with the other requests, it is included in the Collection where we will collect all suggestions (see >Procedures section above).
Summary
In summary, to improve the baseline version significantly we should create a better user design that guides a user in a better way, we should add documentation, and we should decide on the naming/explanation of the fields
“Better user design”: there will be more time allocated for front-end design and more testing for V2.
“Add documentation”: this is constant work in progress. The workflow is: (1) Liliana (with Jap's support) explain how the MS and its recipes actually works draft guidelines here, (2) Improvements based on testing and demonstration scenarios are compiled both in Gitter/Github (for concrete issues) and in the future (major improvements), design documents here, “we should decide on the naming/explanation of the fields”: see >Procedures above (development project created for this).
“Add documentation”: this is constant work in progress. The workflow is: (1) Liliana (with Jap's support) explain how the MS and its recipes actually works draft guidelines here, (2) Improvements based on testing and demonstration scenarios are compiled both in Gitter/Github (for concrete issues) and in the future (major improvements), design documents here, “we should decide on the naming/explanation of the fields”: see >Procedures above (development project created for this).
HERE STARTS NEW DISCUSSION THREAD
This is because in this recipe you don't search, but only analyze the collection. The other recipes are dedicated to search, thus, your selected collections can be sent either to comparative search or to single search (there will be another recipe coming soon for combined search -with no comparative emphasis).
Btw, please add any ideas you may have about this in the Recipe design documents, the requirements specifications for Github issue #9 will be written based on the ideas suggested in those documents.
Well ,I will post it and come back later to see if there is an answer. I get the announcement "No video with supported MIME type found" in the Resource Viewer; I am trying to watch NAMENS …: Kees van Baalen (20-05-1971) (id: 180448@program). Is that a problem that can be fixed?
This is a video from the collection of The Netherlands Institute for Sound and Vision, which is available for play-out at this moment only to the CLARIAH research pilot scholars via VPN. You could also come to the institute where you would be able to play this and other videos.
Hi there, the films with ID 116441@program and 108592@program, both refer to the same film, while they should be different. Could this be checked?
Hi users of the Media Suite, we are trying to evaluate whether the name of this recipe is appropriate. We would like to hear what you think of these options: o Collection inspector (leave as it is)
- Collection metadata inspector Collection metadata analyzer
2.Collection analyzer - Metadata analyzer
- Metadata transparency tool
- Data critique tool
6.Collections
Thank you all for your insights! @tpoell_twitter : Option 6 is a bit overlapping with our data sources service (CKAN), I would say then that the final candidates can be then: a) Collection metadata analyzer, b) Collection metadata inspector, c) collection inspector: a tool for metadata analysis, we will work on more sketches and inform about progress as soon as possible
@RaoulDuke85 Hi Rob, I think this is one of those cases in which you have the same item broadcasted twice. @jblom: could you please help me out in confirming this? : searching by ID 116441 brings result "WELKOM IN HET LEVEN, LIEVE KLEINE (BIS)" with these two dates: 15-01-1964 / 24-01-1982. A query for this title (in quotation marks, withouth BIS), brings the other item (ID 108592) which has the same title with the first "sortdate" only. Thanks!
@RaoulDuke85 Just a kind request: this room is for the Collection inspector recipe, for search-related issues the room Wp5-mediasuite may be more appropriate ;) thanks!
I have new requirements for the Collection Inspector: (1) percentages in the legenda e.g. 14000/28000 (50%) and preferably also relative frequency in the timeline, (2) option to select more than one field to inspect at Y-axis, both cumulatively (take the selected fields together) and separately (for each field another line in the timeline, OR). I assume we will look at this when preparing the new designs for the CS recipe.
Thanks @j_vangorp_twitter. We included this: (1) adding percentages to the stats presentation in the Collection inspector is a new issue (#197), should be easy to do as @jblom said. Adding relative frequencies in the time line requires more work and it is in the list already (#156). (2): this is also requested for the search options (see #152), I added this as a new issue (#198). In terms of priorities, all these are included in the backlog for Version 3.
Also, in the graph that shows the completeness of the selected metadata field it would be useful if you could change the date range, so that you can have a closer view on the pattern for this data (which is not available before 2006).
Apologies for late reply. The fields for ASR are separated, you can see them all when you type ASR in the search box, the most important being "words (in: layer__asr)"
The field "text (in: transcripts)" is supposed to include the teletext (TT888) transcripts, but these were removed from the harvesting endpoint (where the Media Suite collects the data from) because it is public, and this cannot be offered like that due to copyright issues. But these transcripts will be offered via the Media Suite (after authentication) in forthcoming versions, not sure if for V4. This field, now, contains keywords automatically extracted from the teletext (see chat in the Tools room).