Help brainstorming potential categories for a 10,000-image system

10000-image-systems
(Nicholas Mihaila) #1

As I mentioned in another thread, I’ve been thinking about developing a phonetic category-based system as an extension of the Ben system. Each image would correspond to a 4-digit sequence for decimals and some would also include a 12-digit sequence for binary. As the title says, I’m hoping to get some help brainstorming the categories.

1600 of the images are directly carried over from the Ben system, so this leaves 8400 images. These 8400 images are then grouped in categories, each of which consists of 100 images. The structure for the naming is [category] + [individual image name], which is bisyllabic and in the form [consonant][vowel][consonant][vowel].

Here’s what I have so far:

tA:
tE:
tI:
tO: toys
n[vowel]: all empty
moo: movies
ma: machines
me: Metroid
mi:
mo: marine life?
mu: muppets
m[long vowel]: all empty
r[short vowel]: all empty
rA: racing (cars, bikes, etc.)
rE:
rI:
rO: robots
loo:
la: lab supplies
le: electronics
li:
lo:
lu: lunch?
lA:
lE:
lI: lights (related displays, games, etc.)
lO: Lord of the Rings
[j/g]oo: jewelry
[j/g]a: gas-powered
[j/g]e:
[j/g]i: gym equipment
[j/g]o: gardening
[j/g]u:
[j/g]A: games
[j/g]E:
[j/g]I:
[j/g]O:
koo:
ka: candy
ke: chemistry (redundant w/lab supplies)
ki: kitchen appliances
ko: cartoons
ku: kung fu
kA:
kE:
kI:
kO: clothes
[f/v]oo: food
[f/v]a: Family Guy?
[f/v]e: fair
[f/v]i: video games
[f/v]o: fossils (dinosaurs, alive)
[f/v]u: fungus?
[f/v]A: fangs (fictional monsters, predatory animals, etc.)
[f/v]E: vehicles
[f/v]I: flying (various aircraft)
[f/v]O: foes (various enemies, villains, etc.)
boo: Boondocks?
ba: battle (war theme)
be:
bi: building (construction)
bo: botany
bu: bugs
bA: babies (baby theme)
bE: beach
bI: biology
bO:

This is just from my initial brainstorming. Some of my ideas were really bad and will certainly be replaced with something else. All suggestions are welcome. I’m looking for an overall composition of approximately 40% inanimate objects and 60% living objects (people, dinosaurs, cartoons, etc.). As you can see, it’s best to have the category align with the first syllable, but this isn’t essential. It’s more important to have categories that lend themselves to producing 100 unique images. Anyway, I’m really busy with school, but I’ll try to update this list to incorporate suggestions when possible.

For more details on how the system works, you can see my posts here.

I appreciate the help!

-Nick

edit I couldn’t get the tags to work.

2 Likes

(Nicholas Mihaila) #2

So it occurred to me that I’m probably going about this the wrong way. Instead of using phonemes to make categories, I should be starting with the categories and then tweaking their key words (category representations) to fit with the phonemes. This is because the categories themselves are far more important than how they align phonetically with their names. Each category needs to lend itself to 100 unique images, and, roughly speaking, each category needs to be equally specific. For instance, it wouldn’t make sense to have two separate categories for “quadrupeds” and “fungi”.

Anyway, this will be my approach from now on. As before, I’m open to any suggestions. Here’s the list that I just started on (I’ll need 84 equally specific subcategories in total.):

Life/nature:

Quadrupeds, African + desert
Quadrupeds, [everything else]
Primates
Rodents and other small mammals
Flying animals + flightless birds
Aquatic, ocean + freshwater
Aquatic, reefs + tide pools + shore + ocean floor (not deep sea)
Deep-sea creatures + crustaceans
Reptiles
Amphibians
Insects, arachnids, etc.
Microscopic organisms (scaled to a macroscopic level)
Plants/fungi
Dinosaurs, land
Dinosaurs, aquatic
Prehistoric/extinct animals (not dinosaurs)
Natural structures/formations + geology/metals
Pets/domesticated animals + supplies

Technology:

Machines (anything that doesn’t fit into any of the below categories)
Vehicles
Aircraft
Water (boats, subs, etc.)
Combat
Tools
Misc. electronics
Medicine
Gym equipment
Construction equipment
Chemistry (supplies, chemists, etc.)

Culture:

Architecture + artwork
Music (instruments, etc.)
Jewelry + related
Christmas + Easter
Halloween
Sports
Countries
Mythology
Ancient history (Egypt, Maya, etc.)
Jobs (police officer, fire fighter, etc.)
Parties/festivals
Art supplies
Clothes/fashion
People (celebrities, family, friends, professors)
Artifacts/antiques (more object-based than ancient history)
Farming/agriculture + western theme
Religious figures/objects

Science fiction + related:

Aliens + X-men characters (excluding those that distinctly fall in either the hero or villain category) Robots
Machines
Spaceships/aircraft
Superheroes/heroes
Villains
Fantasy/mystical creatures

Entertainment:

Shows, child audience
Cartoons, teen/mature audience
Movies, horror + action
Movies, all other genres
Books (Lord of the Rings, Harry Potter, etc.)
Games + arcade + miniature golf + related
Baby/toddler toys
Toys (not for babies/toddlers)
Video games, Mario-like genre
Video games, mature audience
Rides (like those at a carnival)
Anime, action + related genres
Anime, [all other genres]
Disney, princess theme (princess-themed movies and associated characters)
Disney, [all other genres] + Pixar
Muppets + related
Characters/items from card games (MTG, etc.) + Warhammer

Food/culinary:

Prepared/cooked foods (not candy)
Exotic foods
Candy
Cookware + dining
Fruits/vegetables
Cooking ingredients (that don’t fit into any of the above)
Drinks
Junk food (namely fast food)

Household/work:

Hygiene + makeup + related
Office supplies
Furniture + related
Cleaning supplies (vacuums, sprays, etc.)
Shopping items
Home-repair supplies + related (Home Depot store items, etc.)

edit Current subcategory count: 84 (Got it!)

4 Likes

(Josh Cohen) #3

I had started a category-based system for a 1,000 image system, but didn’t get that far. Here are some of the categories. Some categories were repeated (like tools). I’m not sure if it will be any help, but here’s the list. It probably would have been revised many times if I had continued with it.

  • ACTION_FIGURES-HEROS
  • AEROSPACE
  • ANCIENT-WAR
  • ANIMALS
  • ARCHAEOLOGY
  • ART
  • ARTIFACTS
  • ARTISTS
  • BIOLOGY
  • BIRDS
  • CARTOONS
  • CIRCUS
  • COLORFUL-ANIMALS
  • COMPUTERS
  • CRAFTS-OR-MYTHOLOGY
  • CULTURES
  • MYTHOLOGY-DEMONS
  • DOMESTIC-ANIMALS
  • EGYPT
  • ELECTRONICS
  • ENERGY-OR-OFFICE_SUPPLIES
  • WORLD-MUSIC
  • EXERCISE
  • FLYING_THINGS
  • FOOD
  • FRUITS
  • GODS
  • HISTORY
  • INDIA
  • JAZZ
  • LITERATURE
  • LOTR
  • MACHINES-OR-STARWARS
  • MARINE
  • MATHEMATICS
  • MATH-SCIENCE
  • MEDICINE
  • MILITARY
  • MONEY-AND-ARTIFACTS
  • MONSTERS-TALES
  • MUSIC
  • MUSICIANS
  • MYTHOLOGY
  • MYTHS
  • MYTHS-FABLES
  • OPERA
  • PLANTS
  • CHINA
  • POLITICS
  • PRESIDENTS-LEADERS
  • RELIGION
  • RULERS
  • RUSSIA
  • SCIENTISTS
  • SHARP-THINGS
  • SPORTS
  • STAR_WARS
  • STRUCTURES
  • SUPERHEROS
  • TASTES
  • TOOLS
  • TOOLS-OR-UNUSUAL-RELIGIONS
  • TOOLS_WOOD
  • UTENSILS
  • WORKERS-JOBS
  • ZODIAC
4 Likes

(Nicholas Mihaila) #4

Thank you, Josh! I think it’ll help a lot. It’s just a matter of ensuring that all the basic categories are covered and that they’re broken down in a logical way. It’s just easy to accidentally exclude a useful category. It’s hard to identify what you don’t know isn’t there. :slight_smile:

Once I reach 84 categories I’ll work on the revision process to try to make each unique and equally specific. I think this first step is the most important as it sets the foundation for everything else. Once I have the categories, the fill-out process will lend itself to lists that I can find though online searches, so hopefully it won’t be too terrible.

I think going through a lot of revisions is inevitable. I know that when I made my image-name associations for the Ben system, the revisions were countless. It was the single most time-consuming project of my life. The time input was absolutely obscene.

2 Likes

(Nicholas Mihaila) #5

Update:

Many edits later, I think I’ve finally managed to outline all the major categories. I currently have about 70 subcategories, and I need 84. The next step will be be to expand the current list by subdividing categories until the total is 84 and each category is approximately equally specific. Also, as I mentioned before, I would like the composition to be about 60% animate, 40% inanimate.

2 Likes

(Nicholas Mihaila) #6

Done! I just filled out the 84th subcategory.

Here’s the composition:

Animate: 29 (34.5%)
Inanimate: 32 (38.1%)
Mixed: 23 (27.4%)

So as you can see, in the end it’ll be like this: 34.5% ≤ % animate ≤ 72.6%, which encompasses my targeted 60%, so that’s really good.

I’m going to spend at least another week or so revising things, and then I’m going to make an Excel spreadsheet to begin the fill-out process. Does anybody have any comments? Anything I missed? Does the variety in each subcategory look roughly even?

2 Likes

(Josh Cohen) #7

I haven’t thought much about considering whether images are animate or inanimate. I guess most of my images are animated or anthropomorphized in some way, even if they are objects.

1 Like

(Nicholas Mihaila) #8

It’s an extremely rough guideline. It’s based on the idea that animate-inanimate pairs are easiest to link because they can have an action-object relationship, but many animate images can also act as objects, so the ideal ratio is a little skewed from 50/50. This is just what’s worked best for me. At one point I spent a lot of time experimenting with different ratios.

2 Likes

(Nicholas Mihaila) #9

Update:

I put quite a bit of thought into how I should format the list of images for the fill-out process. I decided to put everything into a single Excel spreadsheet organized according to category. This allows me to quickly search all images to see if an object is already in use. Because the suffixes aren’t unique for each category, I’ll be searching key words in the name of the category itself to navigate around the document. After doing a key-word search, I’ll just have to scroll within a list of 100 items to find what I’m looking for.

Because the prefix (first syllable) has no effect on the fill-out process, I’m holding off on that until the list is complete. Basically, the prefix is phonetically based on a key word that’s associated with each category. Often this is just the category itself (ex: ma → machines), but that’s not always the case.

Here’s a picture of what the document looks like:

A 1 in the sum column means that there’s an image for that particular phoneme. Adding them up lets me track my progress more easily.

1 Like

(Josh Cohen) #10

Looks good. I don’t know if it would make significantly less typing, but if you’re counting the number of rows that don’t have an image in column B, you might be able to use the COUNTA function.

1 Like

(Nicholas Mihaila) #11

Thank you. I’ll check it out. I’m not very well versed with Excel functions to be honest.

0 Likes

(Josh Cohen) #12

I think you can click in cell E1 and then type this in the cell:

=COUNTA(B2:B1000)

Replace B1000 with the number of row you expect the sheet to have. The cell E1 should then be replaced with a dynamic count of all the cells in column B that have data in them.

It might be easier to do it your way though. :slight_smile:

0 Likes

(Nicholas Mihaila) #13

So I got the counta function to work. It makes everything really easy. I also used the counta function to calculate the current composition of animate images and my overall progress (% relative to 10,000).

The fill-out process has been going very smoothly. Despite the category restriction, the freedom in selecting the second consonant has made coming up with images significantly easier than it was for the Ben system. I can tell that the first half will be a breeze to fill out.

Here’s an image from the vehicles category:

ex%202

The basic format is image, short description or key word // image. The 6th column is for notes, which are usually just alternative images. The total images is out of 8,400, which is the balance from those pulled directly from the Ben system.

2 Likes