Hay Kranen created a proof of concept where Commons is searched for pictures that (per standard) use a "depicts" statement.

For some reasons, it is true for "kitimaguru", but if I search for "lamp" (EN) versus "lampe" (FR), or "key" (English) versus "clé" (French), I really do not get the same results at all

Hoi, Florence I totally agree that proper internatonalisation, localisation is key. What is key for me is that this already provides an easy and obvious search function for mediafiles that have a link to a Wikidata item. Just to stress the point, this is a wiki, we do not need a fully functional search engine (for all the Commons files); that is what we aspire to that is what we work towards.. That will take years. But with a proper search tool, a tool that makes it EASY to use Commons, it may fool me into using Commons for my blog. To show you that it works, I just looked for "baisikeli <https://nl.wiktionary.org/wiki/baisikeli#Swahili>" and made a screenshot [1]. The screenshot is with other files showing the evolution of this tool in a Commons category [2] Important to notice is that the tool DOES invite you to localise the labels to French, Swahili et al for best results!! A minor observation, there are all kinds of things that could change in the user interface. Key is that this is a prototype. It is showing us how we can make Commons work for us. Thanks, GerardM [1] https://commons.wikimedia.org/wiki/File:Appelmoes3.png [2] https://commons.wikimedia.org/wiki/Category:Hay%27s_SDSEARCH On Sun, 24 May 2020 at 01:21, Florence Devouard &lt;anthere(a)anthere.org&gt; wrote:

Hi all, It’s worth remembering that this functionality is built in to Commons, it’s just not as user-friendly. From the example below, if you put "haswbstatement:P180=Q191931” into the Commons searchbox, you will get the same results. Thanks to the structured data on commons project+team! Also, around half of the Commons categories now have multilingual labels embedded in them through the Wikidata Infobox, which means that if you do an ordinary search for a phrase in a different language, you should find the correct commons category if it exists. E.g., try searching for “Telescopio Lovell”, or "洛弗尔望远镜". The infobox also has a link at the bottom of it that you can click on to search depicts statements for that category’s topic without having to look up the QID first. Thanks, Mike wrote: we be In is obvious Just to search what Commons screenshot tool how less it. , https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines and https://meta.wikimedia.org/wiki/Wikimedia-l <mailto:wikimedia-l-request@lists.wikimedia.org?subject=unsubscribe> _______________________________________________ Wikimedia-l mailing list, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines and https://meta.wikimedia.org/wiki/Wikimedia-l New messages to: Wikimedia-l(a)lists.wikimedia.org Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l, <mailto:wikimedia-l-request@lists.wikimedia.org?subject=unsubscribe>

First of all, good job. I appreciate that. To be honest, however, some of us did not need a "proof of concept" to know that this was possible, we would have helped to build it years ago if the situation allowed it. it did not, the "social" environment was against it. We could not de facto support the metadata architecture with a bottom-up approach and a top-down strategy was necessary to go forward and, as a result, their quality is limited... so any service based on them will be limited as well for a while. It's a step in the right direction and I wish you all the best... you need me to translate in Italian, fine, I am here. The problem is however that it took more time than necessary to get there and it would be useful to face that. You don't have to explain to many of us what metadata were for in the end and what long-term change they can introduce, we knew that because you can see their use in many other platforms an get an idea how they are supposed to work...even newbies with professional experience got the concept that Wikidata could help Commons, by themselves. So the question is IMHO how this changes the minds of active users who did and do not care even now. Something of course changed over the years... nothing is static. When the Wikidata infobox arrived the comments of distrust of Wikidata drastically reduced, but it was not long time before this introduction that some Commons users were still insulting you for leaving a welcome template on their Wikidata talk page. Is their attitude the core reason why these efforts are late? IMHO it is. Is it going to be solved? Because if it's not, this introduction wlll be in any case very slow. A bumpy road. We will have two mentalities coexisting at the same time, as we have for example with metadata. Also, what does it mean for us who are active also in the real world? Not a greta change yet if the timeline is slow. When I went to a third party I had to say Commons was late on many issues (categorization, metadata, search engine, metrics, copyright guidelines, lack of analytical instrument for the backlog, workflow of NC files...) and no doubt such gaps were going to be filled one day but not soon (definitely much slower than many other issues on other platforms). Now I can say it's still late and is (as expected) catching up on this issue. Still, the third party won't be impressed, the reply will be "good. call me when it's ready". Which is fair, they are doing their job. I can't change that yet. I don't know how my attitude can change it. I still think that this relies on the attitude of the bulk of users that even now are not interested in dealing with such long-term issues of Commons and any effort will wait a lot to get a clear feedback. Il domenica 24 maggio 2020, 15:38:36 CEST, Gerard Meijssen < gerard.meijssen(a)gmail.com&gt; ha scritto: Hoi, Mike you are absolutely right but you are missing the point that I am trying to make. Yes, what is exposed by Hay's SDSEARCH tool is based on all the work that was done before and as such it relies absolutely on the work that has been done before. Without it, this tool would not be possible. This work is key for us to move forward. What is so vitally important about this proof of concept is that it readily opens up Commons depending on a localised user interface. Even when that is not available search, it is possible based to search based on the availability of labels in a language. This proof of concept dramatically shows that nothing more is needed to open up Commons to a multilingual public. This proof of concept is an invitation to adopt this approach and make it available in properly internationalised code as part of a multilingual Commons user interface. It invites people to participate and with some social engineering it the shore that turns the ship in making Commons a much more positive place. Why, because making Commons usable even useful is what we have not done for all the languages but English. When people are happy to use Commons, they are more likely to participate and join its community. So far we could not care less as long as it was used in our own projects. The challenge that I present to you is to make Commons *my goto place* for illustrations for my blog. When you can convince me, you convince the world. Remember our approach is that of a wiki. It does not have to be perfect, it has to empower us to move forward. Thanks, GerardM On Sun, 24 May 2020 at 15:12, Michael Peel &lt;email(a)mikepeel.net&gt; wrote: the bottom in promote our gerard.meijssen(a)gmail.com tool, a in get label get https://tools.wmflabs.org/hay/sdsearch/#q=haswbstatement:P180=Q191931 in it and https://lists.wikimedia.org/mailman/listinfo/wikimedia-l https://lists.wikimedia.org/mailman/listinfo/wikimedia-l, _______________________________________________ Wikimedia-l mailing list, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines and https://meta.wikimedia.org/wiki/Wikimedia-l New messages to: Wikimedia-l(a)lists.wikimedia.org Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l, <mailto:wikimedia-l-request@lists.wikimedia.org?subject=unsubscribe> _______________________________________________ Wikimedia-l mailing list, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines and https://meta.wikimedia.org/wiki/Wikimedia-l New messages to: Wikimedia-l(a)lists.wikimedia.org Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l, <mailto:wikimedia-l-request@lists.wikimedia.org?subject=unsubscribe>

Hi Gerard, I mostly agree with you. However, I disagree with this: it Turning prototype code into production code is never trivial. When you’re writing a prototype, you get to skip all performance and edge case concerns, and you don’t need to integrate it into existing code, you’re just interested in getting something working. I hope (and expect) that the WMF will make improvements to Commons’ multilingual search in the future, but it’s definitely not a “very little” amount of work that needs doing, it’s a year or more worth of developer time. Thanks, Mike _______________________________________________ Wikimedia-l mailing list, guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines and https://meta.wikimedia.org/wiki/Wikimedia-l New messages to: Wikimedia-l(a)lists.wikimedia.org Unsubscribe: https://lists.wikimedia.org/mailman/listinfo/wikimedia-l, <mailto:wikimedia-l-request@lists.wikimedia.org?subject=unsubscribe>

It’s worth remembering that this functionality is built in to Commons, it’s just not as user-friendly. From the example below, if you put "haswbstatement:P180=Q191931” into the Commons searchbox, you will get the same results. Thanks to the structured data on commons project+team!