Finding promising releases for fixing (part 1)

So far these blogs have been about detecting wrong information in releases that people entered into the database, but there are also other errors that are much harder to find, namely detecting if a release is missing information and are incomplete.

Many releases in Discogs are missing information. One of the things we did for a while was track new Spanish that were added, looking at the pictures (if any) and then adding the missing information, such as the depósito legal, or a rights society. But this does not scale, and it is also quite labour intensive, and it is not easy to remember by heart which releases have already been checked. So I want to optimize to success: which releases will likely be the easiest to fix?

I ran a test, where I confined my search to:

1. releases from Spain
2. 7" releases (vinyl, flexi-disc)

with the specific goal to fill in the depósito legal field(s) and other things like rights society fields (if not already there).

The following was done: first I looked at which releases are from Spain. In the dump covering March 1 - 31 2019 there are 298,468 releases from Spain. I then looked at the XML from the dump and discarded all the releases that have no images. Whether or not there is an image is stored in the image as in the following example:

<release id="7146474" status="Accepted">
  <images>
    <image height="602" type="primary" uri="" uri150="" width="600"/>
    <image height="606" type="secondary" uri="" uri150="" width="600"/>
    <image height="603" type="secondary" uri="" uri150="" width="600"/>
    <image height="601" type="secondary" uri="" uri150="" width="600"/>
  </images>

If the images element is empty the entry can be skipped, as there won't be any useful information to extract from it.

I also only looked at the 7" releases, that did not have a Depósito Legal field and also could not be found in the list of smells (indicating a possible or faulty Depósito Legal, although this is not fool proof). The result: 36,008 releases (and remember, this is just 7" releases, no LPs, CDs, or other releases that also have this number), distributed like this:

I then sorted these in reverse order, with the releases with the most images first and started working my way through. As it turns out, this is a very effective method to prioritize releases that need to be fixed, but only as long as there are enough images: when going through just a handful of releases with 4 images most of the releases (my guess: 3 out of 5) had a depósito legal value that could be added. For releases that only come with a single image (a bit less than 18,000) the success rate plummets drastically, as most of the images are simply the front sleeve (and this is why Discogs should start labeling the uploaded images as well).

That being said, not every release I looked at was suitable, because:

• no depósito legal present on the release, because it was either too old, actually from another country, or simply didn't have it
• the pictures did not have enough detail or were very fuzzy
• stickers on labels blocking the depósito legal
• impossible to read data, even with a high enough detailed scan, such as this Randy Newman release

I am very sure that this method can be expanded to other information as well such as Czechoslovak license numbers or Greek license numbers. But before I do that there are 36,000 potential depósito legal numbers to add!