1

Topic: Wikidata integration

Looking to integrate Redump with Wikidata, the inter-language portal and metadata system for Wikipedia.  The Redump.org ID (P5290) was approved today, weeks earlier than expected.

Re: Wikidata integration

Huh?

PX-760A (+30), PX-W4824TA (+98), GSA-H42L (+667), GDR-8164B (+102), SH-D162D (+6), SOHD-167T (+12)

3 (edited by user7 2018-06-12 18:38:24)

Re: Wikidata integration

I'm guessing this is to use Redump as a reference.

Join us for a chat if you can so we can talk more Dispenser:  https://discord.gg/qmY8Dxm

Re: Wikidata integration

Wikidata is essentially a semantic web project, it's a site "similar" in idea to wikipedia (human collaborative curation) but aiming to build a structured knowledge graph.

I don't know what OP is looking to do, but a quick look at the page it seems like he would like to replicate Redump DB's info (some of it, at least) on the KG.


What level of data are you planning to place into the graph?

5

Re: Wikidata integration

My interest is in linking all video games information together.  So a link from Redump.org would take you to Wikidata, from Wikidata you could visit Metacritic, MobyGames, or Giant Bomb.  Or somebody could write a script to pull those links from Wikidata.  Theoretically, it could auto-generate missing entries, but the community would need to drop their "biographies-only" focus and start working on other topics.

There also issues of "re-use" I'd like to address, like the XML .dat dumps have no IDs to get back to the site.

Re: Wikidata integration

Well, keep in mind one thing… The Redump ID is merely a database convenience. The XML files you are looking at (The DAT files) are NOT complete data dumps. most of the data is in fact contained on the site's DB, whereas the .dat file is a special-purpose format for file management software. To have a complete picture of the data, you need at least both sets.

moreover, keep in mind Redump is a media preservation project. Each single ID refers to a single piece of physical media the group has documented (i.e. dumped, verified, and collected metadata for). This means that an ID won't be uniquely associated to a video game, but to a specific disc of a specific edition of said game.

It would be great to meet in chat over on discord as creating a broader KG of videogames is something I think is very intersting, but perhaps needs a little more pre-work than simply starting from the Redump IDs.