User talk:Oravrattas/Archive 2


Deputies are not members of legislative terms edit

Hi. I appreciate your refinement of legislature period, but I'm afraid the way you've chosen is not the correct one. The deputies as persons can be members of various organizations or groups, but not members of items like 7th Czech legislature (Q21140425), which according to it's statements is an instance of time period. Please stop automated adding of these statements and let's try a better way to include this (certainly useful) information into the items. Greetings, --Shlomo (talk) 14:52, 18 October 2015 (UTC)Reply

  • Hi — on this I was following the approach that Andrew Gray has been using at Wikidata:WikiProject British Politicians, so it's important to include him in the discussion too. I think part of the issue here is that in some countries, each legislative term (Q15238777) is treated simply as a time period, whereas others talk about them as the body itself. See, for example, http://www.riigikogu.ee/en/introduction-and-history/history-riigikogu/12th-riigikogu/ — where the 12th Riigikogu is very clearly talked about as actually being the legislative body. It would definitely be good to come up with a clean way to model this. --Oravrattas (talk) 15:16, 18 October 2015 (UTC)Reply
    The approach where every election period is considered a separate body is possible, with some limitations:
    1. It would have to be described as an instance of a legislative body, i.e. subclass of organization.
    2. It's not applicable to all elected bodies; it can be applied on Chamber of Deputies of the Parliament of the Czech Republic (Q320265) (although it's not common to speak about "7th Chamber of Deputies" in Czech language, but let's put linguistic consideration aside and concentrate on the fact, that every new elected chamber starts from zero; it needn't and even mustn't continue the unfinished procedures started by the previous chamber. Though, this is not the case of e.g. Senate of the Czech Republic (Q389423), which acts continuously, having every 2 years 1/3 of it's members new elected. I suppose, other countries have bodies like this too.
    3. There are already (thousands of) statements following another approach and stating that a person is e.g. position held (P39) Member of the Chamber of Deputies of the Parliament of the Czech Republic (Q19803234). If we consider the chamber in every election period a separate body, we actually shouldn't have Member of the Chamber of Deputies of the Parliament of the Czech Republic (Q19803234) as an instance, just as a class, and we should create instance items like "Deputy of the 1st Chamber of Deputies of Czech parliament", "...of the 2nd one", a.s.o.
    4. We have to take in account, that there are also some bodies, where a part of members is elected while some seats are reserved for virilists or nominated or co-opted members. It is questionable, whether these can be described as "members of x-th chamber of something", if their membership is not dependant from the election period of the elected members.
    So even when your (=Andrew Gray's) approach is in principle possible, it requires solid analysis of every particular institution and it's rules to determinate if it's applicable for it. And even if it's possible, it still can be more clever to use the system that is already established and enables a more uniform approach to various type of parliaments. Even if it is theoretically a little bit less precise in some cases.--Shlomo (talk) 16:33, 18 October 2015 (UTC)Reply
I hadn't noticed that legislative term (Q15238777) is defined solely as a period rather than a group. I think changing this restriction would probably be a better way forward - I've not been able to think of a better way to model this than going with P39. Perhaps we could change legislative term (Q15238777) (or just the subsidiary items like parliamentary term in the Kingdom of Great Britain (Q21095053)) to be subclass of (P279):legislative term (Q15238777) and subclass of (P279):group of humans (Q16334295)? Andrew Gray (talk) 15:54, 18 October 2015 (UTC)Reply
I suggest keeping legislative term (Q15238777) as a subclass of time period, so it can be connected not only to the deputies, but also to the documents, proceedings, organizational structures a.s.o. I would suggest to use the existing structure of statements based on position held (P39) and add the legislative term (= an instance of time period) as a qualifier. We'd have to find or propose some appropriate property for this though. But beside of universal (I hope) usability of this approach, it also can solve the problem, how to describe that a politician who was e.g. a deputy in 2nd to 6th election period, was also the chairman of the parliament in 5th and 6th period and a faction leader in 3rd to 5th period.--Shlomo (talk) 16:51, 18 October 2015 (UTC)Reply
I see the logic here... but one problem with using the term as a qualifier on position held (P39) is that we'll also be wanting to add electoral district (P768) qualifiers - in most cases, they were a member of parliament for somewhere. This might make P39 quite complex - it has a single entry with qualifiers for)electoral district (P768), "during" (all the legislative terms), plus possibly start time (P580) and end time (P582) - which don't always match the terms. Andrew Gray (talk) 17:08, 18 October 2015 (UTC)Reply
Well, I suppose, that's what qualifiers are for... I would add even more of them: replaces (P1365) and replaced by (P1366) (where applicable), political party which nominated him to it's ticket (not necessarily the same as member of political party (P102)...), link to the election (or more general: to the event constituing his mandate), maybe more...--Shlomo (talk) 17:56, 18 October 2015 (UTC)Reply
Yes — I think it's valuable to be able to associate things other than the members with a Term. It's also worth noting that it's very common that official websites will list, for all current members, the previous terms they were also part of (e.g. the last column of http://www.riigikogu.ee/tutvustus-ja-ajalugu/riigikogu-ajalugu/x-riigikogu-koosseis/juhatus-ja-liikmed/), and it's useful to be able to capture that information (especially for cases where they don't elsewhere provide the full lists of members of those terms). I don't really have strong opinions on whether that's done by multiple distinct position held (P39) entries, or something else — it's simply important that we can model it. And, of course, many of the things people want to do with this information is on a per-legislative term (Q15238777) basis too — who the members were; how many seats each party/faction had; what legislation was passed; etc. — and especially so if you want to compare things like the make-up over time (by party, gender, whatever.) So ideally it's something that can be queried directly from that, especially for cases where we know that someone was a member of the 3rd and 5th Assembly without having any idea of the dates of those terms. --Oravrattas (talk) 17:21, 18 October 2015 (UTC)Reply
It might also be worth noting that in the Popolo standard, a Term is neither an Organization nor a Time Period, but an "Event", which in turn can have an associated time period and organization. This doesn't solve the direct problem of what it is that the individual politician is a member of, but it's something worth bearing in mind in relation to the wider modelling question, as I think at least some of the confusion here can come from legislative term (Q15238777) being not quite either of those things. --Oravrattas (talk) 17:53, 18 October 2015 (UTC)Reply

Please stop, your tool is marking non-person concepts as humans edit

See this screenshot from your contribs, and another example is [1]Commemoration of Atatürk, Youth and Sports Day (Q5152297) --BurritoBazooka (talk) 18:03, 24 October 2015 (UTC)Reply

I have now blocked you for 6 hours. Please write here when you have seen this message and promise not to add wrong statements anymore, so I can remove the block. Regards, --Stryn (talk) 18:31, 24 October 2015 (UTC)Reply

Many apologies for this. This was entirely my fault, though it took me a while to track down what happened: I was setting 'human' on everyone who was missing it from the Turkish Politicians category. However, the Autolist tool by default descends into sub-categories as well, which I hadn't noticed, so was picking up many more pages that I hadn't checked, and which indeed aren't actually people. I will refrain from making any more edits using that tool without very carefully checking every page that it's going to alter, and will also roll-back any changes that I made in error. --Oravrattas (talk) 19:24, 24 October 2015 (UTC)Reply
Thanks for explanation, I have now unblocked you. I already reverted a bunch of wrong additions, but I think there are still some left. --Stryn (talk) 19:41, 24 October 2015 (UTC)Reply
Thanks. I think I've cleaned them all up now. In as many cases as possible I didn't just revert the previous edit, but changed it to a more suitable value. --Oravrattas (talk) 22:10, 24 October 2015 (UTC)Reply
It's also unclear to me how Autolist deals with categories. A graphical representation would be nice, and there's no indication to the user what "NS" does. What I usually do is go through the list of articles it has found, and check/uncheck the checkboxes next to the article links in Autolist. For instance, to find all the Sumo stables, I used Autolist to find articles in ja:Category:相撲部屋 whose title also contains the character "屋" (autolist results), manually went through a few dozen articles, checked the checkboxes of the ones which were definitely Sumo stables (there were a few about scandals surrounding Sumo stables, etc) and set autolist to run on them. After that, while the first Autolist instances is running, I open a new tab and check the next few dozen, and repeat. Time and patience is saved. --BurritoBazooka (talk) 21:50, 24 October 2015 (UTC)Reply
Yep — that looks like a much better approach! Thanks. --Oravrattas (talk) 22:10, 24 October 2015 (UTC)Reply
Return to the user page of "Oravrattas/Archive 2".