User talk:Hoverfish/Sandbox

The following are points I am gathering in order to focus on starting a project-wide discussion in WP Films on the issue of categorization. This is 1. to become aware, clear out, define, how we as a project see categorization in our area, 2. to see if we agree on some basic lines to clarify in our categorization page, 3. to be able to give consistent advice to new members or sound and clarifying arguments to editors whose work happens to run contrary to ours, and 4. to hopefully improve films categorization to serve better a wider number of users.

What are film categories for?

 * WP: categories are important structures for searching for related articles
 * Let us speculate further.
 * Should we consider them mainly as a tool for us editors more or for the reader?
 * Obviously we, as a project, can come to some consensus about how we see them and how we need them to be.
 * But can we possibly know how the reader sees them and uses them?
 * How far can we guess this without projecting our own needs or tendencies in it?
 * However we can consider what a reader CAN do with them and what NOT.

Considering the various groups of users
What may be the primary function they perform for various groups of users?

Considering the reader
In what ways can a reader use them?
 * Navigation through articles.
 * To find something similar or related.
 * To find lists of interest.
 * Exploration of the categorization tree.
 * Curiosity, education, fun...
 * For search usability.
 * How easy is it for the casual reader to learn about inctegory searches?
 * Even if one learns, does this search work satisfactorily? If no why?
 * Is it because of the limitations of the feature or the way we have been categorizing articles, or both?

Considering the editor
What are the reasons to create them?
 * Classification in meaningful groups.
 * Classification in useful groups.
 * Creating lists and sublists of articles (for classification, for maintenance, for other work)
 * Navigating through various groups (for maintenance, for other work)
 * For the net joy of putting everything in etticketed boxes. > Much of this could be a subconscious impulse that one may not even be able to explain fully, except find various short-range justifications.
 * To augment one's number of edits. > This is bad faith, I know, but I can't help wondering sometimes.

In what ways can an editor use them?
 * Navigating back and forth through various areas of interest or work.
 * To have indexes of articles of an area of interest or work.
 * To get results from the search box of wikipedia or from PetScan.
 * Use them as keywords for the above mentioned searches... (to what extent is this possible?)

Narrowing down to WP Film
Some questions:
 * Should categorization of film articles facilitate searches?
 * Is categorization of film articles there to summarise certain characteristics?
 * or to define certain common characteristics?
 * Should we distinguish between various uses of film categorization?
 * To what extent do we use categories for various aims (maybe make a form so that everyone can fill in one's specific needs and habits)

Good arguments that have been given so far:
 * In support of keeping some basic all-inclusive categories:
 * For search usability. Help:Category
 * For recent changes usability. Help:Category
 * _
 * In support of splitting categories to subcategories
 * To find films that share some common characteristic in one place
 * _

Searching specifics

 * Help Desk question: Help_desk/Archives/2017 July 13... In a nutshell can the incategory search be made to look for categories that contain in their name Term A AND Term B (AND term C, etc)? Wouldn't this be absolutely terrific? Some slight renaming in some subcategories to make sure key terms in the parent stays intact in the subactegories and there you have it! incategory search for ANY category containing "comedy" AND "films" / incategory search for ANY category containing Scorsese!! Categories are suddenly useful and user-friendly for all!


 * From Wikipedia FAQ:
 * Categories FAQ (for readers)
 * Categorization FAQ (for editors).

In the first, the general public is informed that there is a type of categories called Universal categories, used to provide a complete list of articles which are otherwise normally divided into subcategories. In WP Film it has been decided to keep such "Universal categories", irrespective of subcategorization, for the three primary categories of by Year (YYYY films), by Country (CCCC films), and by language (XXXX-language films).

In the second FAQ we are told that there are two main ways to use categories: lists and topics. A usefull point made here is that there are some natural hierarchies of lists (like scientific classification), which leaves it to be understood that not every area falls under such a natural hierarchy categorization. In films we have to articles that could belong to two or even three basic genres. Generating specific categoris for each subset (blach comedy films, romantic comedy films) and removing the article from the parent (comedy films) makes it much more diffucult is not impossible to use categories to search for useful lists, unless ones has the patience to do many specific searches (which also requires knowledge of all the subcategories we have created). Imagine using search to find all the comedy films of Scorsese. What does one do? Aha comedies have been broken down to by country, so incategory:Films directed by Martin Scorsese and incategory:American comedy films... this gives 0 results. So we have to discover the articles have been removed from "American comedy films" and subcategorized as "American black comedy films". Now we get 2 results, but could there also be other subgenres of comedy he has created? Lte's go to his filmography them. OH nuts, the list doesn't mention genre. So we have to get to all his films and see which genre or subgenre categories each one has been assigned. Fail!

Intersecting categories
There has been, and still is, a feature request for Wikipedia: Category intersection. It is partly because of the input from the key person behind this request, that in the early days of Film categorization the idea of keeping three all inclusive basic categories in all film articles was incorporated by consensus in our categorization scheme.

Classification and usability
I got a donation of 50 kilos of mixed screws in a big bag and I have to sort them all so that anyone in this workshop can use them... Let's see... This is how I would arrange my screw storage bins
 * Screws by gauge (tag)
 * various gauges (not very useful for sorting ANY screw, but gauge tag needed in all)
 * Screws by length (tag)
 * various lengths (not very useful for sorting ANY screw, but length tag needed in all)
 * Screws by type (this is my usability preference)
 * Wood screws
 * Wood screws by type of head
 * Wood screws by type of head and gauge
 * Wood screws by type of head and gauge and length (here go the actual screws)
 * Machine screws
 * Machine screws by type of head
 * Machine screws by type of head and gauge
 * Machine screws by type of head and gauge and length (here go the actual screws)
 * Thread cutting machine screws
 * Thread cutting machine screws by type of head
 * etc
 * Sheet metal screws
 * etc

Now in comes a screw collector and points out that have I mixed all the screws of the 19th century with the ones of the 20th and 21st century and he brings in some colleagues who support his view and they start subdiving the screws in separate bins by centrury. Fine, this is fair by consensus. It will take me a bit longer but I can still find what I need. But next comes in a screw manufacturer who insists that screws of different manufacturers have to be placed in different drawers in the bins.

Now films and screws are not things of the same nature. Obviously different criteria both in classification and in usability apply. But the example is good to highlight the question, where does too little or too much classification make a group pointless or useless?

I copy here from this past thread where Bearcat had suggested country-genre crossings for reasons of cross referencing with other corresponing categories. Finally country-genre crossings have been created across films but each film has also been kept in its parent by-country category. The issue for me with categories is that there should be discussion about the "optimum" index level in any taxonomy. As you say, for this taxaonomy it should be "Canadian films". An "optimum" level means that there might be an article that corresponds with the category, and once created, the category should not be broken up. This implies that smaller groupings would either be a list, or if a category, there would be duplicate entries. This is not the norm, (though I've been advocating this for several years now, and I support your efforts here.) -- Samuel Wantman 06:22, 7 April 2007 (UTC)

Comments
Regardless of which approach is taken I think "genre" should be non-diffusing as well along with country, year and language. I think most searches will revolve around patterns like 2009 American films or British horror films, or 2012 action-comedy films etc. If genre were non-diffusing then "incategory" searches could be used to search pretty much any permutation of film combination. If we did that then categorization structure would at least cease being a functional issue and become just an organizational one. Having a non-diffusing comedy category wouldn't preclude inclusion in a sub-category such as Category:1980s parody films but at the moment relegating a film to 1980s parody films from the comedy categories does prevent an effective "incategory" search of those categories. It's not a straight one or the other choice here. I think part of the problem is that a small group of gnomes have decided how these categories should be used, so perhaps it would be a mistake for us to assume how they will be used too. We should look to broadening their use and purpose IMO. Betty Logan (talk) 16:15, 17 July 2017 (UTC)

(forgot to ping) Hoverfish Talk 15:02, 18 July 2017 (UTC)
 * I have done some 180 turns in my views about subdividing big categories in smaller significant ones. My main turn to a middle path was when I realized that the need for BOTH more very specific AND some more generic all-inclusive categories. On the issue of genre I was never clear about what the best way may be, but after you pointed out the category search features I tend to be FOR having some basic all-inclusive genre categories IF possible (see below). But which? And will they stick? And are we ready as a project to face the complains that are bound to come about keeping so many parent categories present along with their kids in films?

Wikipedia_talk:WikiProject Film/Categorization/Archive 1 - My question is, how does the all-encompassing Films category help a user to find a film? The film industry is genre-driven, and most users look for films by their favorite genre, director, or maybe decade or style. If we have a comprehensive Films category, will it actually help anyone except our department? As a librarian, my main concern is to make information more accessible to the user. I have no, repeat, no opinion as to whether we should have a comprehensive Films category, but I think it should be a low priority. My 2¢. Cheers, Her Pegship (tis herself) 18:39, 2 January 2007 (UTC) Wikipedia_talk:WikiProject Film/Categorization - You can use any of the categories to relate a quality of the film. That's not the point. The idea is to have a few specific markers which should be easily definable and non-contentious in and of themselves. (How we choose to define the category standards may be well-debated, but whatever standards emerge should not be subjective itself.) My point is that there is no verifiable way of choosing a genre category. It's not that it's not a useful category, but for the purposes of essential categories which should exist with no excuse, the other listed cat types are fairly easy for anyone to lookup and assign for a given article. Girolamo Savonarola 01:11, 28 March 2007 (UTC)

From then on, and since there was no strong opinion on genre categories, various editors started creating genre-related subcategories until some went too far and there was consensus to limit some, but all I see is the patern "it says so in the guideline" or "no it doesn't" and so on. There was never a clear discussion on genre, so I am all for having one. But first we have to know what exactly we are suggesting. Which genres can be all inclusive for films? Last month Film genre was tagged for need of better citations and possible OR by an editor who tried to improve the refs. In any case it doesn't make anything clear enough to be of help here. Hoverfish Talk 18:00, 17 July 2017 (UTC)

This is why I am so desperate to find where I can discuss about the possibility of having a search syntax that goes like incategory: [term in category name] AND [term in category name] AND [term in category name]. But the help desk archived my question without a hint...Hoverfish Talk 18:10, 17 July 2017 (UTC)