[09:18:42] how can I reorder constraints now? [10:20:52] A country language name explorer using Wikidata https://blog.mapbox.com/exploring-the-world-with-wikidata-and-openstreetmap-30f1bfe954d3 [16:36:38] addshore: available? [17:11:36] MisterSynergy yup [17:13:59] heyho [17:14:08] did you see my request at testwikidata, addshore? [17:16:15] Nope! [17:16:21] Cn you link me? [17:16:39] https://test.wikidata.org/w/index.php?title=User_talk:Addshore&diff=prev&oldid=205153 [17:24:57] {{doing}} [17:24:58] 10[1] 10https://www.wikidata.org/wiki/Template:doing [17:25:33] thnx [17:25:46] MisterSynergy: done [21:44:38] any admins around? [21:49:59] !admin [21:50:14] !admin I have a few questions... [21:50:44] Maybe it doesn't require an admin? [21:51:31] Yes? [21:52:28] Hmm. You are right Trizek, anyone with some knowledge about wikidata and its origins and probably its relation to freebase will do [21:53:05] Please tell us your question then! :) [21:53:35] So, I'm working on a project where my goal is to replace the existing module that uses Wikipedia as the knowledge base with a module that uses Freebase. [21:54:26] AFAIK Freebase has been turned off [21:54:29] Freebase is closed since June 2015. All the content has been transferred to Wikidata. [21:54:31] Definitely there are a lot of differences in the way the freebase data is stored as compared to Wikipedia (Which is basically pages containing links to other pages and descriptive text) [21:54:35] Indeed it has been [21:54:47] Not all data has been imported. Some has. [21:55:06] many Wikidata items have a freenode mapping [21:55:17] Freebase :-) [21:55:20] My question therefore is, how would the comparison between Wikidata and freebase look like [21:55:22] * freebase [21:55:38] hoo you use a lot of irc channels huh :) [21:56:10] I do have access to the freebase data that Google provides online (It is the API that has been deprecated) [21:56:30] So, could you tell me how many entities wikidata contains as of date [21:56:32] roughtly [21:56:48] 31-32M [21:57:00] and is there a mapping between Wikidata's QID and freebase's MID for those entities [21:57:23] Ah, the process has been slow I guess. Freebase has about 1.8 billion entities [21:58:17] Is the Primary sources tool the only way to migrate data from freebase to wikidata [21:58:18] ? [21:59:19] Yes, many entities already have a freebase id associated [21:59:39] https://www.wikidata.org/wiki/Property:P646 is the property [22:01:02] Ah. Thanks nikki [22:01:55] Are you guys familiar with any disambiguation strategies that use freebase or wikidata? [22:01:55] Well, freebase entities are not exactly Wikidata items [22:02:01] but I can't quite recall the details [22:02:25] What do you mean by that? Like I have "Berlin" and I want to find out which Berlin that is? [22:02:47] Exactly! [22:03:32] I expected Wikidata was a subset of freebase. Not in the structural form. But that there is a one-to-one mapping from all (almost) wikidata entities to freebase entities [22:03:43] We have https://www.wikidata.org/wiki/Special:ItemDisambiguation?language=en&label=Berlin but that's very basic [22:03:47] Is that a fair assumption? [22:04:02] Not really… I guess we diverge quite far [22:04:22] especially when it comes to proteins or scientific articles, freebase hasn't had many of these [22:04:29] Pardon me if my questions seem lame, I just want to do my research well about these datasets before choosing one [22:04:37] Ah [22:05:06] Would I be able to find a comparison of wikidata v/s freebase online? in the form of a blogpost or a research paper etc [22:05:22] No idea [22:05:33] maybe [22:06:13] I'll look around. Thanks for now. I'll probably be bothering you guys with more (hopefully meaningful :) ) questions soon [22:06:52] One last question, at what rate is Wikidata expanding? [22:08:03] AdityaAS: one correction to the above already: freebase has 1.9bn *triples*, not entities [22:08:17] wikidata is already larger in that sense [22:08:48] AdityaAS: Well, depends on what we're counting [22:09:09] hoo: Let's say we're counting triples [22:09:12] on most metrics we probably have an exponential growth [22:09:33] hoo: I know you've worked on the property suggester, do you know much about the interface? for a long time I've wanted a way to always suggest certain properties, I wouldn't mind a gadget or something, but I have no idea whether it's technically possible to insert my own suggestions (let alone how) and no idea who to ask [22:09:36] Clearly exponential, usually 0.5-1.5% per week. More currently (due to a mass import of papers) [22:09:58] nikki: It's possible as a gadget, but not quite fun [22:11:03] MisterSynergy: so how many unique QID's does Wikidata have currently? [22:11:05] it can't be less fun than typing the same property names over and over all day :P [22:11:13] I had to stop several times today because my hands hurt too much [22:11:19] 30M and a little more [22:11:37] okay, almost 32M [22:11:43] would you happen to know the number in the case of freebase? [22:12:04] no, they only provide the number of triples on their last condition [22:12:10] which was around 1.9bn [22:12:21] https://developers.google.com/freebase/ [22:12:38] But if I have the data, I should be able to figure out right. I'll try to find that out [22:12:42] There is a chart about this number in Wikidata, but I don't find it right now [22:13:02] it is larger than 3bn IIRC [22:13:05] for Wikidata [22:14:06] That is impressive. Have any recent research papers (published in the last 3-4 years) used the Wikidata knowledge base for any task? [22:14:11] Any papers that you know of [22:14:32] okay, it is 2.5bn apparently [22:14:33] I just want to know if there are others like me, who have already worked with Wikidata [22:14:51] Wikidata only started in 2012, so you wont find anything older really ;) [22:14:56] MisterSynergy: Are the stats posted somewhere online [22:15:05] https://grafana.wikimedia.org/dashboard/db/wikidata-query-service?refresh=1m&panelId=7&fullscreen&orgId=1&from=1470861802973&to=1502489752997 [22:15:16] There have been quite some publications… should be discoverable rather easily [22:17:55] I'll look for them. I don't know much about the information stored about each entity. But is this the info that is stored? https://www.wikidata.org/wiki/Wikidata:Introduction#/media/File:Datamodel_in_Wikidata.svg [22:18:16] yeah [22:19:07] Where could I find the exact field names. Like freebase ID was P646 i think. How would I be able to refer to aliases, descriptions etc. (a link would do) [22:25:19] I hold of a bunch of stuff which I think I need to read before asking further questions. I'll be back [22:25:20] :quit