Yeah, via digging I saw that post, but since the ticket wasn't explicitly for it, I figured it should have its own ticket. I checked out the Trello ticket, I'm glad it's getting some support, but I don't have a trello account set up so I couldn't vote.

Looks like the current use percentage has increased, but is still lacking - 7% movies, 10% TV Shows, 0% People

I'm a dev for Jellyfin and I'm looking to make a plugin for wikidata as a metadata source, once I have that hook connected, it'll be a lot easier for me to update the wikidata database :)

We're adding the Wikidata support on our end in the next month or so. But we still need to see our ID's get matched at a higher rate. Hopefully this will help.

The API is designed to pull only, is there any pushing allowed? One solution is to automate the matches, but to prevents errors, it would still need to be manually reviewed and approved. What's the review process for editing? I've made a few additions and they appear to go live instantly.

If machine edits are a strict no, I can handle it differently on my end by possibly adding a button that'll direct end-users to the tmdb link directly to allow for manual edits. Tad slower but it'd still allow for convenient additions.

@travisbell said:

We're adding the Wikidata support on our end in the next month or so. But we still need to see our ID's get matched at a higher rate. Hopefully this will help.

Since I'm a Wikidata admin since the beginning of the project, let me know if I can help you in some way to speed up the matching.

@travisbell said:

We're adding the Wikidata support on our end in the next month or so. But we still need to see our ID's get matched at a higher rate. Hopefully this will help.

There's been some progress made on the Wikidata side. Might be a good time to review the current match rate. Thanks!

For reference of anyone following or just finding the Wikidata external ID thread - the feature is now live and was added Oct 2022 see: https://trello.com/c/PP0CpPcu/43-add-wikidata-id-support

Do you have any idea how to link TMDB and Wikidata records properly now that we have Wikidata ID support on TMDB? I'm assuming that we should just add Wikidata ID on TMDB when there is TMDB ID on Wikidata end already. But how to link records when there is no TMDB ID on Wikidata end? Should we add both TMDB ID on Wikidata and Wikidata ID on TMDB, or is the Wikidata end going to be synced automatically with TMDB somehow (e.g. based on IMDB IDs added on both ends, or based on Wikidata IDs exposed via TMDB API)?

@talestalker said:

Do you have any idea how to link TMDB and Wikidata records properly now that we have Wikidata ID support on TMDB? I'm assuming that we should just add Wikidata ID on TMDB when there is TMDB ID on Wikidata end already. But how to link records when there is no TMDB ID on Wikidata end? Should we add both TMDB ID on Wikidata and Wikidata ID on TMDB, or is the Wikidata end going to be synced automatically with TMDB somehow (e.g. based on IMDB IDs added on both ends, or based on Wikidata IDs exposed via TMDB API)?

(note: Wikidata admin here)

Well, adding TMDB ID on Wikidata is fairly easy, and you can do it even if you're not logged in / registered, so I would suggest to do it on both ends. A synchro would be great, indeed, but that would require someone that periodically runs a bot on both ends - it's doable, but requires some sort of coordination.

I strongly suggest to NOT create new Wikidata items when there are none on the Wikidata side, at least for the time being.

Thanks for your answer! However, can you elaborate a bit on this:

I strongly suggest to NOT create new Wikidata items when there are none on the Wikidata side, at least for the time being.

What do you mean by creating "new Wikidata items" in relation to TMDB? Do you mean that it is not recommended to add Wikidata ID only on TMDB end with no corresponding TMDB ID added to Wikidata record on Wikidata end? (I have no problem with adding IDs on both ends, just asking.)

Edit: And as for the syncing – I assumed there is already somebody doing that on Wikidata end based on the initial batch of TMDB IDs filled on Wikidata (perhaps based on IMDB IDs?)...

@talestalker said: What do you mean by creating "new Wikidata items" in relation to TMDB? Do you mean that it is not recommended to add Wikidata ID only on TMDB end with no corresponding TMDB ID added to Wikidata record on Wikidata end? (I have no problem with adding IDs on both ends, just asking.)

Sorry, I wasn't clear. Let me try again with an example.

If Anthony Hopkins has an entry on TMDB (https://www.themoviedb.org/person/4173) and on Wikidata (https://www.wikidata.org/wiki/Q65932), if Q65932 has no TMDB ID, you can add it on Wikidata.

If Anthony Hopkins has an entry on TMDB, but NO entry on Wikidata, I suggest NOT to create (yet) an entry on Wikidata, because it's not sure if there is compliance with Wikidata's criteria for eligibility.

Hope it's more clear now. :)

Edit: And as for the syncing – I assumed there is already somebody doing that on Wikidata end based on the initial batch of TMDB IDs filled on Wikidata (perhaps based on IMDB IDs?)...

We have an external tool where you can manually match IDs from TMDB, but nothing that does it periodically. I think it'd be easy to create it though. I can't do it myself, but I think I can help create the request on Wikidata.

Hope it's more clear now. :)

Yep Thanks!

I can't do it myself, but I think I can help create the request on Wikidata.

Imo, it would be great, but I think that @travisbell should have a say in this.

I'm fine with a periodically matching IDs. I'll be getting the API endpoints for exposing the Wikidata ID's shortly, so there will be a proper programmatic way to trying to find these as well.

@Sannita said:

We have an external tool where you can manually match IDs from TMDB, but nothing that does it periodically. I think it'd be easy to create it though. I can't do it myself, but I think I can help create the request on Wikidata.

@Sannita are you referring to Mix-n-match as the existing external tool or another tool? If Mix-n-Match is not the tool suggested, which one are you referring to? (I'm a regular Wikidata contributor, my username: Wolfgang8741 (Q57978392).)

As of my recent review Mix-n-Match catalogs, they are not up to date nor complete and would either need a dump of the existing TMDB IDs to upload for matching to Wikidata or the Mix-n-Match tool would need to be granted an exemption so it would be allowed to scrape for the IDs which would periodically keep the tool up to date with the new TMDB entries. Either would enable quality reports of Wikidata matches. Source of thehis Mix-n-Match tool: https://bitbucket.org/magnusmanske/mixnmatch/src/master/

Once matched in Mix-N-Match it would facilitate matching back to TMDB as well as monitoring Wikidata for deviations based on its catalog reports, though I wouldn't call a match from Mix-n-Match the only check that should be performed. Before syncing a Wikidata ID to TMDB, the ID sync tool should perform additional verification checks on the match and be able to periodically verify any variation of the values after the initial match and sync on either TMDB or Wikidata, more so on Wikidata if TMDB locks the value. An example additional data to check for automated ID sync of a Movie would at be verifying there are matching values for the publication year and director in addition to the title, depending on what would be matched the additional criteria should be set to match through automation.

• TMDB movie ID: https://mix-n-match.toolforge.org/#/catalog/1066 -> https://www.wikidata.org/wiki/Property:P4947 -> reports/queries: https://www.wikidata.org/wiki/Property_talk:P4947
• TMDB person ID: https://mix-n-match.toolforge.org/#/catalog/1067 -> https://www.wikidata.org/wiki/Property:P4985 -> reports/queries: https://www.wikidata.org/wiki/Property_talk:P4985
• TMDB TV series ID: https://mix-n-match.toolforge.org/#/catalog/1068 -> https://www.wikidata.org/wiki/Property:P4983 -> reports/queries: https://www.wikidata.org/wiki/Property_talk:P4983
• TMDB Collection ID aka film series: no mix-n-match catalog -> https://www.wikidata.org/wiki/Property:P11805 -> reports/queries: https://www.wikidata.org/wiki/Property_talk:P11805
• TMDB company ID: https://mix-n-match.toolforge.org/#/catalog/5992 -> https://www.wikidata.org/wiki/Property:P11806 -> reports/queries: https://www.wikidata.org/wiki/Property_talk:P11806

Has anyone started work on a resolver for the IDs between TMDB and Wikidata? I'm potentially interested in testing an ID match and ID sync tool design, but haven't started coding it yet.

If the sync tool hasn't been started, the probable and most important next step to work out and agree upon is the workflow and quality checks that TMDB would find necessary to match a data integrity criteria for matches between IMDB and Wikidata. We can discuss the matching criteria and workflow here or take this to a new thread or the a repository where the experimental the matching application would be written. If I'm to pioneer this, I'll create a Github Repo and create the tool as an external open source app since what I'm thinking of would be applicable to work with any external ID sync rather than be TMDB specific. Though maybe a tool like this exists and I missed it in the Wikidata Tools - https://www.wikidata.org/wiki/Wikidata:Tools

@wolfgang8741 said:

I have just recently discovered WikiData and it's so awesome! Thanks for being a contributor.

Where can I keep up with your developments on this particular TMDB project out of curiosity?

@wolfgang8741 said: @Sannita are you referring to Mix-n-match as the existing external tool or another tool? If Mix-n-Match is not the tool suggested, which one are you referring to? (I'm a regular Wikidata contributor, my username: Wolfgang8741 (Q57978392).)

Yep. I'm referring to that one!