Browsing made better: Micro-genres.

[Scuppers]

So, here's my idea. The current system for browsing comics sucks. IF I want to find a good movie, I can look up reviews, google "good movies", or ask a friend. You can usually find good stuff everywhere, for just about any medium. However, for webcomics, I believe that there are plenty of really good comics out there waiting to be discovered, but there are a LOT of webcomics. So, I have an Idea to make it easier to find those.
Micro-genres. We could create a database with tons and tons of genres inside genres inside genres. This sounds stupid, but hear me out. We could make it so first of all, the reader could narrow down what they want or do not want. They could add things to a list of stuff they don't want. For instance, maybe they want a comic that is fantasy, but they don't want a furry comic, or they don't want one based off of a game/movie/book. They could start to narrow it down. We could have all of the microgenres sorted in different ways. If we wanted to look at the sprite comics based on sonic (or have sonic in them) we could sort the results by popularity. Or popularity by age? Or new and upcoming?
These microgenre groups could also elect people inside their microgenre to be part of a spotlight for the next up genre. So all of the high school romance people could pick a best person for them and submit it to the romance genre. We could also spotlight certain microgenres. Dark High Science Fiction comic spotlight day, mayhaps?

Now, when it comes to the actual micro genres, we have a problem. Everyone hates being listed in a genre, but we could make it flexible. And hopefully I could make the over arching categories broad enough to be always useful.

First three things someone could browse based on would be Art style, Setting and Story. Art style is explanatory, but Setting I mean as the place where (or when, or subject matter) the comic takes place in. For instance, maybe you really want a comic based on college life. There are a lot of different types of stories, art styles, and themes inside those, and college life is not a story or an art style. Same for sci-fi. Space is not a story. It is a place. To go farther down, sci fi settings for micro genres could go down for near future, robot somethings, 'nother galaxy... we could even narrow that down to star wars based or whatever.

We could talk about the micro genres too. What do you think? Also, I hope this is in the right place.

[elrotram]

I believe this belongs in "Suggestions" and less in "The Business of Comics" unless you mean all comics in general and not just, specifically, this site in general. Also, don't they do something like this with an advanced search in comics?

[robybang]

I think the only way for this idea to be viable would be a dedicated website for fans of the comics to post reviews and tags a la Yelp or Amazon recommendations. However, in order for a site like this to be sustainable, it would need a large, active. userbase.

[elrotram]

I think the only way for this idea to be viable would be a dedicated website for fans of the comics to post reviews and tags a la Yelp or Amazon recommendations. However, in order for a site like this to be sustainable, it would need a large, active. userbase.

This.

[eishiya]

Agreed with robybang.


Making an large, inclusive database of genres and sub-genres is a lot of work, and most of them wouldn't be used on this site. Even if the genres are provided by users, I suspect microgenres would only be used once or twice and most people would stick to the most basic genres. Moreover, people would still have to be listed under certain genres to get at certain microgenres that fit their comic, even if the parent genre does not. Moreover, using your "space" example, a person isn't likely to care what "space" is if they want spare stories. They just want to be able to type in "space rocket sexy" and get stories involving space, rockets, and sexy characters (or rocket porn). They will probably assume that stories that deal with space enough to have that as a tag/microgenre/whatever are going to be scifi, so there's little point in classifying space as scifi. Also, a drama story about astronauts on MIR might still be of interest to them, even though that's not sci-fi, but rather historical fiction.

Whether there are microgenres or simply non-user-defined tags, the challenge is to define the list of microgenres/tags/whatever. It's easier said than done, since it has to be detailed enough to make it easy for people to find what they want, but at the same time compact enough that people will actually be able to find all the tags they want.


And I accidentally went off-topic in suggesting a user-tag-based system, so if anyone's interested, here's a bunch of blahblah about tags.

Spoiler! :

We've had a tag-based system suggested multiple times, and I think that it's a better alternative, especially if combined with the other features and the categories of genres/tags you're talking about.

What if it was a simple tagging system? No sub-genres/sub-tags, as those are very difficult to keep track of, and still lock people into genres (what if someone makes a comic with all the trappings of a noir, but with a scifi-ish adventure premise? With subgenres, they'd still have to get listed under Mystery or wherever Noir is usually classified). In a simple system, they could just have a bunch of tags, like "dark, noir, sci-fi, adventure, novel adaptation" (comma-separated tags).

This could be expanded to be more like what you're talking about by classifying each tag. For example, each comic could have two fields for tags: Art and Plot, or something similar.
Art would have tags for the art production methods and style, e.g. "digital, dark, black and white, realistic"
Plot would be about the story and how it was produced, as well as traditional genre classifications, and possibly other important story-related keywords, like types of macguffins features e.g. "noir, furry, dark, novel adaptation, adventure, sci-fi, flying car"
This opens up the possibility to having the same tag describe different things - like "dark" in this example. Some stories have dark stories combined with cute, cartoony art for effect, and some have both - and being able to search for such stories would be great. But even with a single field of keywords, this could work since most of them are exclusive to either art or story concerns.

Of course, this also means people would have to use tags properly, and... well... the reason most genre systems where authors pick their own genres tend to be simple is because people can't be trusted to use the system properly. The reason a site like Netflix can afford to have a huge array of genres, mood/plot keywords, etc is because they are assigned or at least validated by Netflix's team.
Websites like the Boorus have issues with redundant and inaccurate tags, and the only reason that their tags aren't completely useless is because everyone with an account can edit and modify them, which isn't something a comic author would probably want for their artwork (on the Boorus, art is generally not uploaded by its creators, but by people with only as much investment in it as anybody else). On the other hand, Pixiv has a similar system of tags, but people can add tags (which can be removed by the author).

When a user/author enters tags to add to a comic or search by, the system should pop up some suggestions (similar to how google suggests searches based on what you've typed in so far; this is fairly simple to implement). This guides the user so that they don't end up searching for tags that authors haven't thought of. Tags that are used by only one comic (or maybe less than 3) should be purged from the system on a every year or so to keep the database clean. Purging too frequently would make it too difficult for new tags to take root. Together, these two features (suggesting tags upon adding/searching and purging underused tags) allow the system to self-regulate with minimal human moderation. Unfortunately, they also only work if people actually use the tagging system.
The auto-purging can be done away with if a human moderator regularly checks the tag list (or rather, just the list of those tags used by < 5 comics) and removes any tags that aren't likely to be searched for or used by other authors, such as "drawn by Lana" and "buy male enhancement".


Honestly, I wonder how many people would use such a system to its full extent on SJ. I think when people want to find new webcomics to read, I would imagine they would to want to look for comics outside of one site. A better system for SJ never hurts, of course.

Also, moving this to Suggestions.

[Scuppers]

I think that your sub tags thing is perfect! That is what I'm going for! You see, there would be different sorters, or different classification systems:

Art
Setting
Story

Inside those we could break it down! Art could be simple at first: Just stuff that is easy to classify. So I'll pick manga. I don't mean that manga is easy to classify, just that manga is cleary a different art style than others. Most people that are searching manga don't care about the different micro genres inside manga. But then they could go and narrow down the settings to a "Real World" category, then maybe even more narrow, a crime setting. I would have a drastically smaller search, and I would get things that are interesting. If nothing else, at least I could narrow it down by only getting manga, rather than being presented with a list of just updated, comic spot light, etc.

In truth, instead of offering one big list of genres that are broken into more genres, I would like to add multiple classification systems. These would help shorten each other.
Heres where the Microgenres might help. If I decided to read... Mokepon (So sorry that I'm using this, I really love the comic) then I could see all the comics like them. I could see all the manga- pokemon based- adventure comics, or any other combo of those things. I could pick pokemon/manga, adventure manga, kanto pokemon - adventure, scaling back and forth...

Also, users would hopefully list their own comics. By adding themselves to genres, they are only increasing exposure.
Let's say user A and B have the same sonic based sprite comic. A lists his as a sprite comic, while B lists his as a sprite comic > fan comic> Sonic comic. A and B both get hits for sprite comic searches, but B gets all the additional tiers down.
I know this might be a taken advantage of system, any extra good ideas? Like maybe a if users flag a guy more than three times for being innaccurate type thing? Though that could lead to abuse.

Also, I could post a draft of my classification ideas.

[elrotram]

I also feel like having everything broken down like a dropped glass pan shattering on the floor will lead to more inaccuracies than better representation. Look at deviantart.com - they have sub-genre upon sub-genre and yet, most of the artists couldn't stumble into the right genre with a road map. I feel that, with a basic knowledge of search-engines and a little viewer discretion, readers could discover all the things your micro-genres would offer but on their own. Besides, in their search for "The One" other comics would gain more exposure. But what do I know?

[Scuppers]

I see. That is a good point. It's true that with too many and too much classification the system becomes useless. However, at the moment, all the what, 41 thousand comics scattered across SmackJeeves are hard to find unless you are told the name by a friend, see it in a just updated, or see it in a forum post type thing. Of the thousands of comics, most are concentrated at a very small percentage of this community.
You're right, a search engine is a great idea. However, it is not made for browsing. I don't want to make this like a library. If I want to find anything, anything at all, I just need to look it up. However, the idea in browsing is to spread the viewers interest in thousands of different hidden options. For every Hunger games in the library with 112 holds, there are quality books that are being passed over by most. Even if I wanted to find a good webcomic, the way I would try to find it in a search engine would be by googling "Top webcomics" or "Best Webcomics." I would want someone else to browse for me. A search engine is great if you know what you want, but its not great for those comics which are hidden and the reader who isn't looking for anything in particular.

While it's true that readers could find more comics with patience and hard work, it isn't easy, convenient, or something you would do without thinking about. I think if we made it easy like in a bookstore to just meander through and see all that the store has to offer there would be more exposure to all comics, and increased interest.

Of course, this assumes that there are good comics that are not visible to most viewers. I don't know how many comics there are that are good, but I assume there are enough to make this worthwhile. I thought of this idea because I think that it is hard to find a lot of the good comics out there. If you don't make it to most popular, I imagine it takes a lot of self promotion to make it on the page before!

Also, I think it would help if the categories were broken into easy to delineate boundaries, and have categories where there couldn't be crossover. For instance, in Deviant art they have a Traditional art category, manga/anime category, and fan art. These are all hard to distinguish. What if I'm making a painting of toon link?

[eishiya]

Unfortunately, any sort of categorization is going to hide good comics a person might be interested in, it's just a part of the concept. An even bigger problem is that most of the time, people don't actually know exactly what they want (even if they think they do).

Sadly, no one has thought of an elegant solution to get word out about good, under-appreciated comics yet. I think the best the artist can do is advertise, and the best readers can do is promote the comic via worth of mouth (that's probably the best way, as people will tend to share the comic with people of similar taste, people who will probably like it too).

Perhaps if we try to make a browsing system that approximates the way word-of-mouth works, that would work. SJ tried to do this with the "Similar Comics" feature, but it's based only on the percentage of shared fans (and time of last update), not that great. It's hard, if even possible, for a computer system to figure out what people would consider recommending to like-minded people. Even Favs aren't representative, since a lot of people +Fav comics for very specific reasons (such as it having an uncommon/unpopular kind of art they like, or it's a friend's comic they feel obligated to follow) and likely wouldn't recommend those comics to friends.
How would that a pseudo-worth-of-mouth system work?

Spoiler! :

The simplest way I can think of is that every user can opt to "recommend (comic X they're reading) to fans of (comic Y in their fav list)". The database would simply have a table of (recommending user, comic X, comic Y) tuples (all of which are keys), the recommending user field is just there to allow people to edit/remove their recommendations. The recommendations would be bidirectional, meaning that if I recommend X to fans of Y, then that's the same as Y being recommended to fans of X. Then, the comics with the most recommends with the ocmic you're looking at (perhaps weighted inversely to the number of fans, or expressed as a percentage of fans, so that popular comics do not dominate) are considered the most similar.

Unfortunately, this makes for a big extra table, and it also means more work on the users to actually do this and make it work (the current Similars calculation could be used to fill gaps in recommendations).

Oh, and in case you're wondering how this system helps browsing when it's just replacing the Similars System, which is limited to a comic's profile: A new search feature should be added that would allow something like SimilarTo:ComicID, or perhaps "include similar comics in results" which would find comics that match a (presumably specific) keyword search, and then also include comics similar to those in the results. If there's a "SimilarTo:ComicID" sort of search, then the comic's profile should include a link to that search, as should each comic's mini-profile that shows up in the search. Or something. Searching isn't as easy to program well as it sounds.

[elrotram]

Oh, are we allowed to add metaTags to our webcomics?

[eishiya]

Oh, are we allowed to add metaTags to our webcomics?

Yes. You can add HTML meta tags in your overall template. If you want something searchable by SJ, you can add a list of tags in your comic's description (or if you want to be creative, use those tags in sentences in your description, it makes no difference to SJ, but looks nicer).

HTML meta tags aren't very useful as the biggest search engines don't even use them (or use them in a very minor capacity), but there's no harm in adding them.

[mikemacdee]

I dunno about all of this, but I would at least like "western" added to the genre list. That's what my comic is, really; and that's what it's categorized under on Drunkduck. Calling it "action" or "crime/mystery/thriller" isn't accurate at all.

[Troll]

^
Your option is to label it as "American".

Western is a better term IMO.

[eishiya]

^
Your option is to label it as "American".

Western is a better term IMO.

"American" is in the style section, it has nothing to do with the genre, and will help no one find the comic better if what they want is a western.