Survey: How many labels you usually assign to the photo?

[vlad]

Hi Mike,

do you have the "Also assign its parents" option enabled in your labels?

No. I also don't embed synonyms in the image file.

Synonyms have no effect on the number of labels anyway - but they indeed may affect the number of final keywords. As a matter of fact, it would be interesting to run a similar script on the number of final keywords in the metadata (which may be larger due to synonyms and to enabling "Include all parent level labels as keywords").

You're correct that I use the David Reicks controlled vocabulary. The reason I'm so compulsive about assigning catalog labels is that I enjoy making theme-based slide shows.

That's a very good reason.

As an example, If I want to show photos of circular red objects, I can easily make that happen, not that I've ever done that.

Actually, inviting family & friends to show them a circular-red-themed slideshow seems like a very cool (and artsy) idea. Believe me: I'm only half-joking here!

I'm still waiting for more reports. vlad, how about to participate in survey?

Hi Andrey, thanks for the invitation - but I'm going to decline for now, since my current results would not be useful (even to me):
1) I have many imported images which are not yet cataloged at all.
2) I don't (extensively) label all the images that I import, review and archive - only certain selections (collections). (I know, that's weird.)
3) I'm considering some labelling strategy changes (aka: forever experimenting )

This said, let me tell you kudos for writing such a cool script and promptly updating the graphs, it's interesting to look at them! From what I gather, the label number distribution for most people is close to normal (gaussian) and peaking somewhere around 3 to 6 labels, followed by a long tail to the right. (Actually, I think that's called a log-normal distribution.) It's funny, because the only exception so far seems to be Hert, which has a completely different graph: it shows a logarithmic distribution, with a vast majority of single-label images and then exponentially decreasing percentages as the number of labels gets higher. Just wondering: is this a case in point of less is more? (I largely agree with this principle myself - I just find it hard to apply it in practice! )

A diagram with a single synthetic graph, displaying the overall label number distribution based on all the individual distributions (percentages), would be very useful too!

[Hondor]

From what I gather, the label number distribution for most people is close to normal (gaussian) and peaking somewhere around 3 to 6 labels, followed by a long tail to the right.

I think my own distribution is easily explained. I only make (or keep) people photos, so practically all my photos have exactly one place label plus as many person labels as there are persons in the photo.

People excelling in photos of inanimate objects would probably need some combination of several labels in order to be able to retrieve specific photos from a huge collection of (to my impression) basically the same types of motifs.

[Mike Buckley]

Synonyms have no effect on the number of labels anyway - but they indeed may affect the number of final keywords.

In my earliest days of cataloging, I was using synonyms and embedding them as keywords in the image files. There was also a combination of settings I was using (I don't remember what they were) that automatically read the file and added the synonym keywords as assigned catalog labels. So, depending on other settings it is possible that the synonyms become assigned catalog labels.

I stopped doing that when I realized how many synonyms the David Reicks Controlled Vocabulary has. (I use it.) Almost every word in it has synonyms listed.

All of those synonyms were removed from the image files and as assigned catalog labels when I finally switched from writing delimited keywords to writing hierarchical keywords to the image files as part of some fairly substantial housekeeping that I did before migrating from IDimager to Photo Supreme.

By the way, when I do get around to creating the slide show of red, circular objects, I'll be sure to invite you as a featured guest to the premiere viewing.

[bimo]

Hi Andrey,
my contribution:

Label_Stats_20141101.pdf

(111.96 KiB) Downloaded 844 times

Michael

[Andrey_Ra]

tstoddard thank you, and excuse me for a delay

Tstoddard's Collection

Frequency_tstoddard.png (69.4 KiB) Viewed 16170 times

bimo - such a huge number of labels! Very intriguing. Could you please provide more info about style of labeling that you use?

Bimo's Collection

Frequency_bimo.png (68.33 KiB) Viewed 16170 times

[Mike Buckley]

bimo - such a huge number of labels!

Uh oh. If only one prize is awarded, it could be reasonably argued now that I won't be winning it.

[Hert]

With these numbers, I wouldn't be surprised to hear that bimo is a (micro) stock photographer. Such label counts are typical for stock photographers.

[bimo]

Could you please provide more info about style of labeling that you use?

Shure. My main topics are wildlife and travel photography.

Labels for both kinds: IPTC genre/subject code/scene code/digital source type (4 labels); color indicator (1 l.); location created (5 l.); event (2 l.)
wildlife: species (mostly 3 l.)
travel: location shown (5 l.)
More/special labels might apply where needed (eg persons).

98% of the lables are written to the photo's metadata; but to the keywords section only if no appropriate metadata field is available (ie no location info, no person info etc in the keywords section)

After all it's no big deal with PSU...

HTH Michael

[vlad]

In my earliest days of cataloging, I was using synonyms and embedding them as keywords in the image files. There was also a combination of settings I was using (I don't remember what they were) that automatically read the file and added the synonym keywords as assigned catalog labels. So, depending on other settings it is possible that the synonyms become assigned catalog labels.

You are right: with the appropriate (default?) keyword processing preference set in the "Read settings", new labels could be indeed created for synonyms (upon reading again the synonym keywords from the updated metadata). That's actually something to keep in mind, since it may manifest in practice as an obscure side-effect of certain settings and workflows rather than intended behaviour from user's part! Here's one more reason I'm not going to store synonyms either - as I don't think I'd want them turned to labels in any scenario.

(Ah, so many subtleties - it almost feels like Hert could introduce a PSU training & certification programme. )

All of those synonyms were removed from the image files and as assigned catalog labels

I'd say that is your own use case for less-is-more!

By the way, when I do get around to creating the slide show of red, circular objects, I'll be sure to invite you as a featured guest to the premiere viewing.

Thanks - I'd be honoured!

Hi Andrey, my contribution:

...and finally we have a graph that is almost symmetrical to Hert's graph! (...too bad bimo's statistics don't tell us the maximum number of labels he's ever assigned to a single image - btw, that could be a different contest! )

(Doh, I see Andrey already displays the average distribution; I must have been (yellow-)blind, sorry for the superfluous suggestion in my previous post.)

[Ruhell]

Late to the party as always, but here's my numbers..

Rune

Info
----
Frequency of number of assigned Labels for selected 87138 images:
0 Labels: 7411 ( 9%)
1 Labels: 3035 ( 3%)
2 Labels: 11160 (13%)
3 Labels: 9322 (11%)
4 Labels: 12480 (14%)
5 Labels: 11521 (13%)
6 Labels: 9146 (10%)
7 Labels: 7946 ( 9%)
8 Labels: 6031 ( 7%)
9 Labels: 3654 ( 4%)
10 Labels: 2195 ( 3%)
11 Labels: 1307 ( 1%)
12 Labels: 762 ( 1%)
13 Labels: 440 ( 1%)
14 Labels: 246 ( 0%)
15 Labels: 92 ( 0%)
16+ Labels: 390 ( 0%)

[Andrey_Ra]

Ruhell, thank you for your contribution!
Here is your personal graph:

Ruhell's collection

Frequency_Ruhell.png (73.85 KiB) Viewed 15849 times

[NassauBoy]

Quick question. That script looks like 100% Delphi code.
Just curious ...

[fbungarz]

A bit late, but here my data:
----
Frequency of number of assigned Labels for selected 56613 images:
0 Labels: 14225 (25%)
1 Labels: 4683 ( 8%)
2 Labels: 6340 (11%)
3 Labels: 5170 ( 9%)
4 Labels: 4188 ( 7%)
5 Labels: 5967 (11%)
6 Labels: 4123 ( 7%)
7 Labels: 3243 ( 6%)
8 Labels: 2280 ( 4%)
9 Labels: 1445 ( 3%)
10 Labels: 985 ( 2%)
11 Labels: 1044 ( 2%)
12 Labels: 781 ( 1%)
13 Labels: 576 ( 1%)
14 Labels: 524 ( 1%)
15 Labels: 375 ( 1%)
16+ Labels: 664 ( 1%)

A bit strange is that my catalog count says the catalog database includes 103,892 images, whereas the script says "selected 56,613 images". That's a difference of 47,279 images, which the script does not count, almost half of the images in the catalog. Most of my images are versioned, so perhaps the script does not take versions into account? Else these statistics would be plain wrong...

[freudenthaler]

[ external image ]

Just 4 the records here are my statistics...