Jump to content

Wikipedia talk:Categorization

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia
This is an old revision of this page, as edited by Rick Block (talk | contribs) at 14:56, 1 October 2005 (→Category:Japan geography stubs: make is automatable, run by WP:WSS). The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.


Merging Categories

Is there a way to suggest a category merge like

for articles? I noticed there is both a Category:Eurodance and a Category:Eurodance musicians. 05:03, 17 September 2005 (UTC)

Because categories can't usefully be redirected, we usually take this straight to WP:CFD and suggest deleting one of them. Actually, though, in this case, shouldn't Category:Eurodance musicians be a subcategory of Category:Eurodance? -- Jmabel | Talk 04:51, 18 September 2005 (UTC)[reply]

Discussion page mysteriously truncated

What happened to all previous discussion? Where have the links to archived material gone? Even the straw poll below refers to "Please read the section above", but the section has been removed. -- Jmabel | Talk 17:24, August 7, 2005 (UTC)

Whoa, I have no idea... —Tarnas 21:44, 7 August 2005 (UTC)[reply]
Archived old discussion. Everything's fixed, I hope... —Tarnas 06:43, 8 August 2005 (UTC)[reply]

An article with the same name as a category should usually belong only to that category?

Are there exceptions to the following rule?

For instance, many film, book, and author/artist articles have dedicated topic categories (e.g., Category:Neon Genesis Evangelion), yet it makes sense to have both the main article and the topic category appear in a given larger category (for instance, it makes sense to me to have both Neon Genesis Evangelion and Category:Neon Genesis Evangelion in Category:Anime). Category pages seem like they're built exactly for this purpose: they contain a pool of subcategories in their top half, and a pool of relevant topic articles in the bottom half, and so it seems natural to me that, for instance, Category:Anime would have both the catch-all subcat Category:Neon Genesis Evangelion in its top half and a reference to the specific main article Neon Genesis Evangelion in its bottom half.

It just looks like the topic article under the current policy (Neon Genesis Evangelion, for example, again) lacks categorization at first glance: as they are now, topic articles with topic categories are categorized mostly according to how their topic categories are categorized, and it seems confusing to create that dependency between the two. Was what I'm talking about doing seen as too redundant, and that's why it isn't advocated, or can film articles, for example, be special cases? —Tarnas 23:20, 5 August 2005 (UTC)[reply]

This has been a long ongoing topic of discussion. It seems to me that there are some very good reasons to ENCOURAGE putting an article in both the category with the same name and the category above it. The double listing makes perfect sense to me.
  1. The double category listing sends the message to the user that there is an article about the topic, and there are also more articles to be found in the subcategory.
  2. It makes it easier to find main topic articles (by eliminating having to go to the subcategory)
  3. It puts both articles in the list of categories for the article. There would be the normal higher level listing that all other articles get put into now when they don't have their own category. There would also be the subdirectory with the same name which is the natural jumping off point that people will look for if they want more information about the subject.
  4. It makes more intuitive sense for editors. The natural inclination is to put things into categories, not to take them out of categories.
  5. It creates a complete listing of articles at the higher level category. It is confusing to have to look at the subdirectories and the list of articles to get a complete list of topics.
As an alternative if people want to be really consistent about the no duplication rule, a better way to do it would be to only put an article in the HIGHER level subdirectory. Then both the article and the subdirectory would be listed. In the lower level category, the one with the same name as the article, there would be a link which would be entered manually (e.g. "This topic is about motion pictures. For an introduction to the topic see film".) You would also have to do the same thing in the film article (e.g. "For more film topics see: Category:Film). This would actually probably be a more consistent and user-friendly way to do things than have duplicates, but harder to set up. Either system would be an improvement to just putting the article in the category with the same name. I would like to add some text to the project page about this if there is consensus. Could we take a straw poll to see where people stand? I've added one below. -- Samuel Wantman 00:10, 6 August 2005 (UTC)[reply]
Whoa! I just read through most of this dicussion page, and realized that my question here has been debated A LOT! One theory put forth above is that there's a disconnect between the impulse to categorize and to classify, but I think the biggest problem of all is that there's two competing notions of hierarchy: one that there is a "tree", in which it's possible to have "sub-categories", and the other that there is a web of interconnected articles and categories. This is noted on Wikipedia:Categorization ("Categories do not form a tree"), but the whole Wikipedia:Categorization policy page seems to be riddled with uncertainty about how to apply the notion of a relatively tree-less hierarchy.
     The overwhelming problem, I think, is that there is this concept of "sub-category": there should be no true "subordinate category" when the hierarchy of categories is not a strict tree, there can only be related categories, some more or less subordinate to others. This becomes most counter-intuitive when the idea of "subordinate category" leaks into "subordinate article", where the policy becomes one of including an article only in one category: Deism only in Category:Deism rather than in Category:Religious faiths, traditions, and movements as well. This policy ignores the dual nature of categorization: to organize both related categories and related articles in readily available fashion, not to nest articles deep in hidden categories. To dictate that a given article should not be further categorized is to miss a strong point of relating a topic to others directly on its page.
     The impulse to manage and prune categories has become too counter-productive in this way: there is, for instance, room for all African-American actors in Category:Film actors, Category:American actors, and Category:African-American actors. I think redundancy is okay and desired: it makes things easier to find and reflects the true multi-categorizable nature of most topics. —Tarnas 01:54, 6 August 2005 (UTC)[reply]
Yes it has been debated quite a bit. I can't say we've reached a consensus. I agree with what you are saying (I made many of the same points higher in the page), and I'm trying to keep an open mind and come up with creative solutions. -- Samuel Wantman 07:33, 6 August 2005 (UTC)[reply]

Topic article straw poll

Please read the section above, and respond below. This is non-binding, just a starting place for the discussion to see where people stand. Thanks -- Samuel Wantman 00:10, 6 August 2005 (UTC)[reply]

How should an article be categorized if it is the topic article of a category? Add your name to any of the following options that you find acceptable (Short comments welcome):

A. The article should only be in the category with the same name. That category will be in the higher level category and will appear as a subdirectory. For an example see Film

B. The article should be in both the category with the same name and the higher level category. The category with the same name will also be in the higher level category. Both the article and the category will be listed in the higher level category. This seems to be the current norm. For an example see Musical theatre

  • Easy and functional -- Samuel Wantman 00:10, 6 August 2005 (UTC)[reply]
  • The most natural solution, already widely used. —Tarnas 00:45, 6 August 2005 (UTC)[reply]
  • Jmabel | Talk 17:26, August 7, 2005 (UTC)
  • Easiest from the user perspective. Otherwise some articles of a type will be in a main category while others will only be in a subcategory. Provides a bit of useful redundancy, IMO older≠wiser 22:49, August 8, 2005 (UTC)
  • This is pretty much the current standard and I find it the easiest to use. -Aranel ("Sarah") 23:19, 8 August 2005 (UTC)[reply]
  • Classifies the article in its proper place in the upper category, links the article with the (sub)category of instances of what the article is about, and is not a redundant categorization so long as the article is separated conceptually, denoted by a unique section (" "). See comment below. - Centrx 21:26, 9 August 2005 (UTC)[reply]

C. The article should only be in the higher level category. The category with the same name will also appear as a sub-category in the higher level category. There should be a link entered in the category with the same name directing users to the topic article. There should be a link in the article directing users to the topic category. For an example see Suspension bridge

D. The Wikipedia software should be changed so that articles and categories are merged. There would be space in each article that showed links to "subarticles", and putting [[Category:Panthera]] in Tiger would put a link to Tiger in the subarticle section of Panthera, as well as showing Panthera as a superarticle in the Tiger article (just like categories are now). Some articles would be mostly subarticle links.

  • Best. —Ashley Y 21:23, August 7, 2005 (UTC)
  • Acceptable. This is actually possible now. All it would require is to put the entire topic article into the category page. The disadvantage of this scheme is when you are just looking for the subarticles.

Discussion

  • Plan C. would be best, I think, if we're thinking of the same thing, and I don't think it would be hard to implement. It's consistent with the idea of categories including brief self-descriptions: in the description, the main topic article could be mentioned/linked, and thereby omitted from the actual category. —Tarnas 00:47, 6 August 2005 (UTC)[reply]
    • I think you and I are thinking of the same thing. Yes it isn't that hard to implement and it exists already on many categories and articles. The more I think about it the more I think there are advantages. It does make the notion of what should be in a category clearer, and it might get more people looking in categories. I have come across many new users who don't understand what categories are and how they work. If all the main topic articles had text that said "For more articles about about this subject see: Category:Foo", that helps makes it very clear. Category:Suspension bridges is a good example of how a category would look (notice that Suspension bridge is not be listed under articles.) Take a look at Suspension bridge and notice that there is a link and description to Category:Suspension bridges listed under "See also" but Category:Bridges is the only category the article is in. -- Samuel Wantman 02:00, 6 August 2005 (UTC)[reply]
  • Well, rereading your idea, we aren't thinking of the same thing: since it's impossible put a topic cat in a topic article without the topic article appearing in the topic cat listing (e.g., can't categorize Hunter S. Thompson as Category:Hunter S. Thompson and somehow leave the former out of the latter), the precedent is, I think, to re-list the topic cat so that it appears as the first article in the topic cat listing, under an asterisk or something similar (e.g., [[Category:Hunter S. Thompson|*]]). I think it's nice to actually have the topic cat in the topic article at the bottom, and in the history of wikipedia categorization I think cats used to be at the top, but the format became clumbsy and got in the way of immediate information. —Tarnas 02:08, 6 August 2005 (UTC)[reply]
  • It sounds like you are describing plan B. It is possible to put a topic category in a topic without the topic article appearing in the topic cat listing. You just put it in the article like so: [[:Category:Hunter S. Thompson]] Better yet if it was made the first item in the See Also section:
See all the articles in the Hunter S. Thopson category...
Perhaps I'm not understanding you. I don't want to move where categories display. I still want them at the bottom except for the topic article which would only be in the top section in plan C. Take a look at Category:Hunter S. Thompson. It now looks like it would in plan A and B. In plan C the only change would be that Hunter S. Thompson would not be in the category listings on the bottom, but it would be linked as it is now, on the top of the page. -- Samuel Wantman 07:07, 6 August 2005 (UTC)[reply]
Plan C would remove the HST cat from the blue-boxed bottom of the HST topic article. I'm not into that, in which case I'm for plan B. Having one (very relevant) cat exist in the "See also" section fragments the whole unity of the category-article structure and requires users to learn a new process: topic cats don't show up on topic article bottoms, but somewhere in the end linking text. Not something I like so much. —Tarnas 18:34, 6 August 2005 (UTC)[reply]

Plan B is troublesome because there are often several higher-level categories, and inevitably they get out of sync. It also clutters up the pages of those categories with items appearing as both articles and subcategories. We have a sensible simple rule of an article not being in a category and its supercat, why make a special exception? —Ashley Y 21:14, August 7, 2005 (UTC)

  • That "sensible simple rule" is the exception. The way it is, a topic article can't take on the categories that are attributed to its topic category, so a topic article (probably the most important article in a topic category) isn't allowed to be categorized directly, but must instead be categorized through its topic cat: see Neon Genesis Evangelion and Category:Neon Genesis Evangelion, where the topic article isn't allowed to be (logically) categorized as Category:Anime because its topic cat is. In every other case, an article, not having a topic cat especially about it, is allowed to be directly categorized wherever it is seen fit. —Tarnas 21:41, 7 August 2005 (UTC)[reply]
Sure, Neon Genesis Evangelion belongs to Category:Anime, but only through Category:Neon Genesis Evangelion, so it doesn't directly belong to Category:Anime anymore than it directly belongs to Category:Animation. Putting it in only in Category:Neon Genesis Evangelion is correctly directly categorising it. —Ashley Y 22:04, August 7, 2005 (UTC)
My point is that this rule of indirectly categorizing a topic article through its topic category is silly and not sensible: yes it follows the current rule, but that's what I'm debating, the current rule doesn't make sense to me. I can't see the harm at all in having an article and its topic category listed in the same category, it's slightly redundant but also much more comprehensive and promotes the purpose of categorizing things in the first place: making information easily accessible from poolings of topics. I can clearly see the harm in this current policy: it makes categorizing a topic article that has a topic cat unduly confusing, and obscures the direct link between a relevant category and a relevant topic article by forcing a user who's looking, say, at Category:Anime to pass through Category:Anime series and then Category:Neon Genesis Evangelion before reaching Neon Genesis Evangelion. This extra running around is against the spirit of the dual categorization model: categorizing both categories and articles. —Tarnas 00:17, 8 August 2005 (UTC)[reply]
Are you arguing that Neon Genesis Evangelion should be categorised in Category:Anime or merely in Category:Anime series (as well as Category:Neon Genesis Evangelion)? If the former, then you end up filling Category:Anime with all the series, which seems to defeat the point of Category:Anime series. Categories tend to fill up with articles easily, and creating subcategories is a way of mitigating that, but only if we don't have articles in the supercats as well. —Ashley Y 02:19, August 8, 2005 (UTC)
That's exactly what I'm arguing. I don't think this defeats the purpose of narrower categories, though: just like I was explaining in the "An article with the same name" section above, the only reason to use the current system of socking away articles in highly specialized subcategories is to reduce redundancy, but the whole idea of having a web of related articles interconnected by a non-tree-like category hierarchy is that there should be a certain amount of redundancy: it should be easy to find things, and thus they should usually exist in several places at once (that is, most topics can be categorized in many ways).

As I said in the above talk section, the categorization policy is riddled with uncertainty about the non-tree-like hierarchy axiom: the entire concept of "subcategories" is highly flawed, given that in a non-tree-like hierachy there will be few truly vertically "subordinate" categories, much less "subordinate" articles which are made to only appear in one category. This practice thwarts a primary purpose of the dual categorization model, as I say above: articles and categories belong in more categories, not less, so as to promote more opportunities for finding topical information through category browsing.

Having Neon Genesis Evangelion in Category:Anime, Category:Anime series, and Category:Neon Genesis Evangelion definitely doesn't defeat the point of having the narrower categories: the narrower categories are just that, more sharply defined categorizations of topics—topics which should be covered by varying degrees of precision (that is, in this case, in all three categories mentioned). One thing, though, that we still need to discuss is this idea of cats "filling up". —Tarnas 06:36, 8 August 2005 (UTC)[reply]
Part of the point of narrower categories is to clean up the supercats, so they're not full of subordinate articles, and putting Neon Genesis Evangelion in Category:Anime defeats this. Otherwise, why not put it in all its ancestors, Category:Animation, Category:Film, Category:Japan, etc., etc.? You seem to have no clear rule as to when an article should be in both a category and a supercat. I think it's simplest that they should never be. —Ashley Y 06:44, August 8, 2005 (UTC)
Well, so this is the question of categories "filling up". I don't see much reason not to put an article in all of its "vertical" ancestors, so in the case at hand, Category:Animation and maybe then Category:Art. So what if there are thousands of articles listed in a cat, that provides broad-base access to all relevant articles for alphabetical eye-browsing while there's a means to sift through the articles (choose subcats) at the top of the cat: the article listings will slim down as more narrow subcats are opened up.

But, I'm beginning to see your point. My way of doing things would lead to a bloated tree-like cat hierarchy. Maybe the redundancy I desire in the cat system is being addressed in other ways, like with portals. —Tarnas 07:25, 8 August 2005 (UTC)[reply]
I think there is a separate issue being brought up here. I don't think putting articles in more than one place in the hierarchy leads to them being everywhere in the hierarchy. The problem as I see it is that users expect articles to be listed with their siblings. If their siblings appear in different places in the hierarchy, it seems to make sense that ALL the siblings that fit the categories definition would be there. Thus, the duplication between Category:American actors and Category:African-American actors. -- Samuel Wantman 07:55, 8 August 2005 (UTC)[reply]

Plan C makes no sense to me. Suspension bridge should obviously be in Category:Suspension bridges, it's the most relevant article to the category (though don't change it while it's an example). —Ashley Y 21:14, August 7, 2005 (UTC)

On the surface it might not seem to make sense, probably because it is very different from what we have been doing, but I think it actually makes more sense than the way things are. I'll explain. If you look at Category:Bridges, the top section lists all the subdirectories. Each subdirectory contains articles about individual bridges of each type. The bottom section contains articles about all the different types of bridges and lists of bridges. Likewise, Category:Suspension bridges contains articles about specific suspension bridges. The article Suspension bridge belongs in Category:Bridges but does it belong in Category:Suspension bridges? It is not an article about a specific suspension bridge, it is an article about a type of bridge, and that is why it was put in Category:Bridges. Luckily, we have the very top section of categories which is very much under utilized. This is the logical place to insert a one paragraph intro to the subject and a link to the topic article. If the topic article is linked from the intro, there is no need for it to be linked in the bottom section. The only thing missing is a link from the article to the category. I believe this should be done in bold from the SEE ALSO section of the article (for an example see Suspension bridge.) The reason why I think this is an improvement is because we can explain what is in the category (i.e. "for articles about specific suspension bridges see Category:Suspension bridges"). Plan A and B confuse the categorization scheme. Plan A moves articles away from all its siblings and puts it with its children. Plan B has the article with both siblings and children which makes the relationships harder to understand. Plan C makes the rules of categorization clear -- no exceptions necessary. Keeps siblings with siblings and has explained links to help the user navigate. -- Samuel Wantman 07:17, 8 August 2005 (UTC)[reply]
Well, ignoring A or B, C would be bad for two reasons. First, Category:Suspension bridges is misleading: there are basically only articles on specific bridges in it, and no articles on suspension bridge topics in general except the one list. Following your logic, Category:Bridges would only include articles on specific bridges, since Category:Suspension bridges would only include articles on specific suspension bridges... but of course it doesn't and shouldn't. Category:Subatomic particles might be more representative, if less disciplined: it includes articles on specific subatomic particles, on facilities for observing them, and on theories/discussions of them in different groupings. And second, there's the reason I gave above, which still stands: Having one (very relevant) cat exist in the "See also" section fragments the whole unity of the category-article structure and requires users to learn a new process: that topic cats won't show up on topic article bottoms [in blue boxes], but somewhere in the end linking text. Not something I like so much. —Tarnas 07:51, 8 August 2005 (UTC)[reply]
I grant that your second point has merit. There is the disadvantage you mention, but I can also see that as an advantage. However for your first point, I am not claiming that Bridges would only have articles about specific bridges, I am saying that there are distinctions made about every category about what belongs in the category and what doesn't. Once we make those distinctions (hopefully with an explanation on the category page), If any articles are listed, I'd like ALL the sibling articles that meet the criteria to be listed. Category:Suspension bridges is like many other pluralized categories (e.g. Category:Musicals.) I would be happy with either B, or C (and even D). B and C are actually very similar. B will have duplicate links to the topic article, where C will need a manual link from the article to the category. Perhaps B is better for this reason, it is easier to overlook a duplication than to make a manual link. I could also see a combination of B and C. B would be the case for most categories, but C would be the case for pluralized categories like actors, musicals, and suspension bridges (BTW, the only reason Category:Bridges is singular is because there is also Category:Bridge about the card game.) -- Samuel Wantman 08:16, 8 August 2005 (UTC)[reply]
  • An important reason to include the article in the category with the same name is so that a person reading that article is linked to the category of things that are instances of what the article is about. This would suggest A or B. The problem mentioned above with A is that a user looking at the upper category would not see the article in the article section; although the article can be found in the category with the same name, the article belongs in the category of which class it is a member. So, this leads to B. That B results in the same article in both a category and its subcategory is avoided if the article in the subcategory of the same name is conceptually classified and denoted as the eponymous article which defines the category. Conceptually, this makes the article not part of the category, but the content associated with the category. This has been denoted, in the past at least, by putting the article in its own section of the category, labelled "*" or " " or otherwise. The blank section seems best, for it puts the article at the very top without a section header ("A", "B", etc.). So, B is the best solution, at least in this respect, though the article in the subcategory should be in a section of its own. - Centrx 21:23, 9 August 2005 (UTC)[reply]
  • A combination of B and D, which would require a technical solution, would be for categories to automatically put the first paragraph of the topic article, denoted by " " categorization in the above space that is currently editable. There would be a second editable space for categorization, interwiki links, etc. - Centrx 20:51, 10 August 2005 (UTC)[reply]

Multiple Category Search/Selection method?

One of the problems I've seen with categories is that many are created which are not proper classifications of information, that is "overcategorization", which if the principle is followed leads to endlessly convoluted and redundant category schemes (see Wikipedia:Categories_for_deletion/Log/2005_August_3#Category:Free_Linux_software). Yet, these categorizations are helpful for searching for information that satisfies multiple classifications. So, while "Free Linux software" or "Windows compression software" is not a valid category, it is useful to be able to list or search for Linux software that is Free software and Data compression software for Windows. The way to do this using proper categorization is to have a system by which categories can be selected/searched, so that a person can select a list of articles that are in more than one category, say "Free software" and "Linux software". So, I was wondering if there is any technical initiative to do this, whether under way, etc. or other suggestions. - Centrx 15:27, 10 August 2005 (UTC)[reply]

I certainly support this idea. I am setting up a Wiki and do not want to go down the path of a huge number of categories which become difficult for novices to comprehend. The capacity to do a multi-category search/listing would be great. Can be done using the Search feature now but is not practical as I would like to include simple links in the article that list articles matching more than one category. Thomas 19 August 2005

Fiction

I was wondering about the policy of categoring articles on fiction. I was looking at the articles in Category:Fictional companies and found out out that some of them are also categoried in the non-fiction categories. For example Oceanic Airlines is also in Category:Airlines and Cyberdyne Systems Corporation is in Category:Companies of Japan. Should they be removed from them? --Laisak 13:06, 17 August 2005 (UTC)[reply]

I think so. -Sean Curtin 05:45, August 18, 2005 (UTC)
I would say they ought to be removed. - Centrx 17:34, 18 August 2005 (UTC)[reply]

Updating the section on category duplications.

I would like to remove the following section:

An article should not be in both a category and its subcategory, for example Microsoft Office is in Category:Microsoft software, so should not also be in Category:Software. An article with the same name as a category should usually belong only to that category, for instance, Deism belongs only in Category:Deism.

and replace it with this:

Normally, an article should not be in both a category and its subcategory, for example Golden Gate Bridge is in Category:Suspension bridges, so it should not also be in Category:Bridges. However there are occasions when this guideline can and should be ignored. Duplication is warranted when the set of subcategories are incomplete, when an article is the topic article for a category, and in a few special cases when it just makes common sense. For more about this see Wikipedia:Categorization/Categories and subcategories

I've written the subpage Wikipedia:Categorization/Categories and subcategories. I'd be thrilled if we could finally reach conseunsus about this. We've talked about it forever. I've tried to come up with what I think is the accepted norm in practice, and the consensus of our discussions to date. -- Samuel Wantman 10:11, 18 August 2005 (UTC)[reply]

Hm... It's been 3 days and nobody has said a word. In standard consensus decision making "silence equals agreement". If this goes a week without comment, I'll just make the change. -- Samuel Wantman 08:45, 21 August 2005 (UTC)[reply]
I like the new policy subpage and the new rule system, it needs some tweaking but it's a very good start in a new direction. Keep it! —Tarnas 19:02, 21 August 2005 (UTC)[reply]
  • I'm sorry but no reaction in three days does not equal a consensus (especially on a relatively backwater page like this - try CFD/talk). I'll go and read it now. Radiant_>|< 09:34, August 23, 2005 (UTC)
I fairly strongly support keeping the original wording, as these rules make the category system much easier to use. For instance at George W. Bush if you want to get to related articles you have to search through a fairly large, and ugly, block of text before you find Category:George W. Bush. By contrast Albert Einstein is much easier to use and better looking with the parent cats still easily accessible from the category page. - SimonP 15:35, August 23, 2005 (UTC)
It is a trade-off. You force everyone to do two page loads to get to a parent category. The Bush example is an unusually large list of categories. Another way to handle this situation would be for the first link under "See also" to be something like this:
This would be a very clear way for people to find the category. -- Samuel Wantman 07:57, 24 August 2005 (UTC)[reply]
I support the renaming. I have come across many cases where only one of two subcategories that an article should be placed in exists. We should be flexible. CalJW 02:39, 25 September 2005 (UTC)[reply]

Category Question

I created a category but it says that "There is 1 subcategory to this category" the thing is that it links back to itself. What did I do wrong?? The page is here: [[Category:Historic Houses of North Carolina]]

There was a category link on the page. Fixed it to a better suited category. --Laisak 20:49, 30 August 2005 (UTC)[reply]

Semantic structure

Is'nt this problem a problem of semantic structuring? Lists and categories have something in common - they comprise of a list of items which are hierachically linked.

So what is the difference?

  • Categories usually do not link items by an order other then alphabetical (it's possible though, but it is a pain) Lists primarily do that
  • the formatting of lists gives mor freedom
  • Categories live bottom up, lists top down (though categories can do something that lists can)
  • Pointers versus Edges - Categories use edges, while lists only use pointers. The concept of edges implies automatic handling of "the second side" of the link. (a new category is created only with a new article -> the category page has to be maintained) - the other way around it is not possible (currently).
  • orphaned links can be created by lists, but not by categories (currently)
  • readers often want to know the semantics of a lemma and browse through equivalents or antonyms - tat's why navigation bars are so popular

What we need

What we actually want to achieve is a semantic structure, where we try to use non primitive constructions, without having the primitives integrated yet (Neither categories, lists or navifgation bars are). We need some basic primitives to link articles in a structured mannor.

  • External structuring information: (already done with categories) "element of" <-> "category"
  • Properties: "name of property: value" (currently not well solved) - to make automatic structured list generation possible as well as defining proper prede- and successors
  • list formatter: The lists associated with categories should be formatable by specifying a template
  • navigator: a navigation bar, that displays some prede- and successors and the category, also with an adjustable template (format probably like the category entry + template link)
  • semantic browser: a browser that can read the data and display it like the visible thesaurus or freemind would be really great
  • structuring: It is already demonstrated in Wikipedia:Wikispecies, well we do not have to go that far but the goal would be something like OWL for semantics

--BoP 22:31, 30 August 2005 (UTC)[reply]

Comments

I've read this three times, and I still don't understand what it says. -- Samuel Wantman 05:30, 17 September 2005 (UTC)[reply]

P-) - To make a long story short: If articles can be an element of a category (i.e. article:George Washington element of category:people) they are supposed to have certain properties (in this case: date_of_birth, deceased, nationality, profession etc.). These properties can be used to generate dynamic lists or sort them within groups. (i.e. all presidents of the United States sorted by their date of birth).
These lists would be generated by a template which can be part of a different article. This template might also be a navigation bar taking the reader to the previous or next president of the US. You might also have a real dynamic list of all presidents sorted by any column by choice of the user. Even dynamic charts would be possible. --BoP 17:24, 17 September 2005 (UTC)[reply]
It just isn't going to happen on a consistent basis. It is the nature of Wikipedia articles to vary in completeness. The fact that people aren't instructed that they must do this and they must do that or their efforts aren't up to standard is one of the reasons why so many are willing to contribute. Also, the idea seems to imply that categories are a form of search tool, but one of their greatest merits is that they facilitate random browsing. CalJW 02:44, 25 September 2005 (UTC)[reply]
Categories should be able to assist searching. But these are really software issues, before they can be policy issues. ··gracefool |☺ 15:06, 25 September 2005 (UTC)[reply]

Merging main articles in categories into their respective category pages

I think it would be more efficient use of wikispace if the text in the "starred article" in a category, which summarizes that particular category, went onto the category page per se, and then the article in the main namespace simply redirected to the category page. However, I will probably get significant opposition to this proposal because it would look silly having the name of the main article appear after "Category:" at the top of the page, applying this to all such articles would take a heck of a lot of time to accomplish, and it simply doesn't fit the fundamental style of an encyclopedia. If that is the case, then such an action could only be applied to less-frequently viewed "starred articles".  Denelson83  02:48, 11 September 2005 (UTC)[reply]

But not all category pages are organised this way. Some have more than one article under the * heading. Some have none. And furthermore, the category page (the bit above the list of articles and categories populating a category) is used very inconsistently, with some people using it to merge articles as you say (an example is Category:Nature), and some people using it (correctly) to describe what the category contains, how it relates to other categories and articles, and what NOT to put in the category (examples are Category:Operas and Category:Disasters). I've also seen people putting links to categories in the "see also" sections of articles. Also, your proposal effectively turns categories into articles with a much larger "see also" section. This may be appropriate for some categories. I have seen some categories which, by bringing together related articles, are showing people the potential for a new article about the whole category. Effectively prompting people to write * articles that don't yet exist for a category. This is more appropriate for 'topic' categories than 'list' categories (for which the * article is typically a list). Carcharoth 04:29, 12 September 2005 (UTC)[reply]
Even if all categories had one main article, which most don't, the articles would often be way too long. They subcategories and categorised articles would become footnotes. And I don't want to see the main article for categories I visit frequently every single time I do so. CalJW 02:47, 25 September 2005 (UTC)[reply]

Contra categorization

I posted what was below to express agreement with Larry Sanger's comments; now I see those are no longer here.

I agree. On a number of occasions I've had to explain the seemingly obvious fact that topics lists are far superior to categories, and I've been horrified a couple of times when someone proposed deleting a topics list because there's a category. In particular:
  • Topics lists can readily be organized; categories cannot be organized (except to the extent to which subcategories do that). See, for example, list of probability topics, list of geometry topics, list of combinatorics topics.
  • Topics lists can contain invisible links to talk pages, so that when you click on "related changes", recent edits to talk pages are included. Here's how you do that:
[[omphalology]] [[Talk:omphalology| ]]
Since only a blank space appears after the vertical slash, the link to the talk page is invisible.; it can't be clicked on, but it affects "related changes".
  • An inappropriately named category cannot be moved to another name except by editing all pages that link to it!
  • One can put red links into lists but not into categories.
  • more to come ... watch this space
Michael Hardy 19:37, 23 September 2005 (UTC)[reply]
it is one of the great merits of Wikipedia is that it allows people to use whichever of several methods of navigation they prefer. Many people value categorisation, and where there are both categories and lists the categories are usually more complete now even though the lists had a three year head start, which strongly indicates that categories are more valued overall. If you don't like them, just ignore them. Good lists are also valuable of course, but there are many many poor and incomplete ones. CalJW 02:50, 25 September 2005 (UTC)[reply]

Category:Japan geography stubs

Regarding [[Category:Japan geography stubs]]

With around 4000 articles in this one category, to say it needs subcategories is an understatement. I have proposed two ideas on the talk page, but I figure I might recieve more input if I ask here as well.

  • Split by prefecture
    • Would create 47 subcategories
    • There would only be about 80 articles per subcategory, if divided evenly
    • Would be easier to organize, as the prefecture is almost always mentioned in the current articles
    • Would keep number of articles to a manageable amount
  • Split by region
    • Would create between 8-11 subcategories
    • Ambiguity over location of some prefectures in relation to regions (Niigata Prefecture, Shizuoka Prefecture) would cause problems in categorization
    • Still leaves around 400-500 articles per subcategory, if divided evenly

If it seems okay, then please comment, because I will create the subcategories as soon as I am able to once I recieve some support. I, personally, would prefer dividing by prefectures, but again, I would like some sort of consensus before I even go anywhere.

-Nameneko 01:10, 28 September 2005 (UTC)[reply]

  • What about by type of geographic feature? E.g.
  1. Prefecture
  2. District
  3. City
  4. Town
  5. Village
  6. Mountain or mountain range, river, other topographic feature
  • Would there be any merit in that? District, city, town and village would bear the brunt, and if necessary, could be further detailed by region, e.g. tohoku-village-stub. Or, district-city-town-village could be consolidated and broken up by region, e.g. tohoku-muni-stub. Or tohoku-localgov-stub. Fg2 07:31, 29 September 2005 (UTC)[reply]


Thanks for the speedy feedback! I, personally, prefer the idea of grouping the cities together if we were to follow that idea. My problem with sorting by region is the abiguity and the possibility that there will still be too many articles in the category. I thought of a possible, alternate idea using some of your ideas:
  • [prefecture]-geo-stub or [prefecture]-stub (x47)
  • japan-topo-stub (x1)
However, looking at that, the topographical stub seems almost redundant. Perhaps a category (rather than a stub) could be made for Japanese topographical features, but I think that having a prefecture-per-stub idea would work best for the large amount of categorizing work that will need to be done if/when the subcategories are created.
My main concern is this: is it appropriate to create 47 new subcategories to divide a single category (albeit with 4,000 articles)?
-Nameneko 06:24, 29 September 2005 (UTC)[reply]
Isn't there a way to organize entries within categories? I know the |* alphabetizes before A, but additionally I think I've seen categories with an additional level of organization. If it's possible, why not create categories for regions, and within those region categories, organize by prefecture, alphabetically within the prefecture. That's far below 47. (But I don't know how to do that!) Fg2 07:31, 29 September 2005 (UTC)[reply]
In regards to re-alphabetizing the 4000+ stubs, something like |Miyagi:Cityname or |Aomori:Townname on entries in the Tohoku section for example, would probably do the trick. There's probably a more elegant solution though.
I personally favor adding the prefecture-stub to the entries. Then, having the prefecture stubs be nested sub-categories of each region (Tohoku.Aomori, Tohoku.Iwate, etc.) would solve the 47-subcategory problem. Ambiguity should not be a problem either, because the Niigata-stub sub-category could be in both the Tohoku category and the Kanto-stub category.Neier 14:15, 1 October 2005 (UTC)[reply]
Adding the prefecture name to the stub category sounds reasonable to me, and I think could largely be done automatically using user:pearle or user:whobot. Whatever is decided should be run by the folks at Wikipedia:WikiProject Stub sorting. -- Rick Block (talk) 14:56, 1 October 2005 (UTC)[reply]

Pages without category

I think it is better to start a Category for Pages without Category, as an temp place for Pages without Category before they can be placed in a proper category.

The person who wrote the above MAY HAVE been intending it as a comment on the Japan discussion, in which case an extra "=" at each end of the heading would help. BUT MAYBE that person has never found Special:Uncategorizedpages? Robin Patterson 01:26, 29 September 2005 (UTC)[reply]
Retrieved from "https://en.wikipedia.org/w/index.php?title=Wikipedia_talk:Categorization&oldid=24484030"