The GÉANT OER Metadata Analytics accepts only ZIP files that contain XMLs(internal zip structure doesn't matter).
The maximum ZIP file size is currently set to 2.4 MB
Every uploaded ZIP file and its content are kept to the GÉANT OER Metadata Analytics server for 30 mins then they are deleted.The same applies to the analysis results files(in case you've chosen to get the results as CSV)].So after uploading your ZIP file,you have 30 mins to configure and run the statistical analysis.After the metadata analysis has finished you have 30 mins to download the results(in case of CSV option).
Analysis Consfiguration
When choosing this option you get a ZIP file containing the results in CSV format.If you want GRNet Metadata Analytics to draw the results in charts,just leave this option unchecked.
When choosing this option you activate the calculation of entropy for each one of the elements.This option is not by default added to each one of the analysis experiments since it is a very time consuming process.
If you've chosen specific elements to analyze but they don't use the specific attribute then you won't get any results.The statistical analysis is element driven.
Since the statistical analysis is element driven, in order to do a vocabulary analysis you should also do a full element analysis.
You just need to uncheck all available elements at the elements panel,so just press the Reset button after you have chosen the desired elements for vocabulary analysis.
Analysis Metrics
  • Element Frequency:The number of times an element is found to the whole record set.
  • Element Completeness:Depicts the percentage of a specific element’s usage to the whole record set.
  • Element Importance:The product of element frequency and element completeness
  • Element Dimensionality:The maximum occurences of a specific element to a metadat record through the whole record set.
  • Element Entropy(optional):The amount of information a specific elements carries.The bigger the entropy the more informative an element is.

For more info refer to:GLOBE Metadata Analysis

In essence the attribute analysis is a vocabulary based analysis over a specific attribute.
The frequencies of the various element possible values.
  • Number of records:The size of the whole record set.
  • Average File Size(bytes).
  • Approximate Storage requirements(bytes).
  • Repository InformativenessThe record set average informativeness.Present only when entropy calculation is activated.