Possibly misleading lodstats

[csarven]

There are a number of shortcomings (perhaps misleads) with the stats
that are given for the LOD Cloud dumps at stats.lod2.eu. One obvious
one is that it
takes the first dump URL that's available from the data hub API.
See:

http://stats.lod2.eu/rdfdocs/485

and

http://datahub.io/dataset/world-bank-linked-data

see also:

On 05/17/2012 01:13 PM, Sarven Capadisli wrote:
On 12-05-14 08:08 PM, Sarven Capadisli wrote:

Hmm.. isn't it possible to iterate through all of the available dump
entries? Although the dumps are in separate gzip files, the contents
within are only in RDF/XML. So, shouldn't your script pick-up on all
of the dumps? I'm unclear on this part. :)

I previously opted not to simply iterate through all available dump
entries as many datasets I encountered while developing lodstats had
excerpts/sample data, metadata or mirrors listed with the same formats
and only a few datasets seemed to have multiple relevant dumps like
yours has.
I'll have a look at this issue again!

If you do get around to updating, please ping me :) I'd love to take a
look at it.

I was just thinking about this again, it might be good to point out some
information and stats with respect to this at stats.lod2.eu (or on the
next TRs). Just a suggestion on mentioning lodstat's current scope /
evaluation:

Given that only the first dump URL is currently retrieved from datasets
with multiple data dumps, what percentage of the data or data dumps are
not accounted for?