2.5 Million tags!
LibraryThing just hit 1.8 million books and 2.5 million tags. Since we're going to hit 2 million books soon, I'll talk about the tags today.
As the tags accumulate, they are also generating a lot more value. Tags are mostly useful personally and statistically. Tags are often played up baselessly—as if a few scattered and general tags are of any use to anyone. For statistical purposes you need a LOT of tags, so frequency patterns can emerge and anomalous entries fade into the background. And tags are primarily interesting in concert, not by themselves. Because tags are non-heirarchical and often short, they lack the "context" of something like the Library of Congress subject headings. Other tags can provide that context.
That's why the "tag similarity" algorithm takes many tags into account, favoring recommendations that match on more than one. Take the messy example of a mid-level book, T. E. Lawrence's Seven Pillars of Wisdom. What the heck is that? Its all over the map—literature, WWI, Middle East, Ottoman Empire, Arabia, history, autobiography, memoir, etc. The recommendations try hitting many of these tags at once—books like Fromkin's A Peace to End All Peace (WWI, Middle East, history, Ottoman Empire, etc.) and Robert Grave's Goodbye to All That (literature, memoir, WWI). It's not perfect—Edward Said's memoir!—but it's a hell of a lot better than any single tag could produce.
And, most importantly, every book and tag makes the statistics better.
Lastly, I wonder how LibraryThing's 2.5 million compares. I'm sure Flickr and Delicious have many times that number. But what else is out there? Amazon has encouraged product tagging for about three months, and they have thousands of times the traffic. I wonder how well that's going?
As the tags accumulate, they are also generating a lot more value. Tags are mostly useful personally and statistically. Tags are often played up baselessly—as if a few scattered and general tags are of any use to anyone. For statistical purposes you need a LOT of tags, so frequency patterns can emerge and anomalous entries fade into the background. And tags are primarily interesting in concert, not by themselves. Because tags are non-heirarchical and often short, they lack the "context" of something like the Library of Congress subject headings. Other tags can provide that context.
That's why the "tag similarity" algorithm takes many tags into account, favoring recommendations that match on more than one. Take the messy example of a mid-level book, T. E. Lawrence's Seven Pillars of Wisdom. What the heck is that? Its all over the map—literature, WWI, Middle East, Ottoman Empire, Arabia, history, autobiography, memoir, etc. The recommendations try hitting many of these tags at once—books like Fromkin's A Peace to End All Peace (WWI, Middle East, history, Ottoman Empire, etc.) and Robert Grave's Goodbye to All That (literature, memoir, WWI). It's not perfect—Edward Said's memoir!—but it's a hell of a lot better than any single tag could produce.
And, most importantly, every book and tag makes the statistics better.
Lastly, I wonder how LibraryThing's 2.5 million compares. I'm sure Flickr and Delicious have many times that number. But what else is out there? Amazon has encouraged product tagging for about three months, and they have thousands of times the traffic. I wonder how well that's going?
posted by Tim at 4:45 AM
18 Comments:
Hi :-)
Detail: when I go to the list of all tags (by having the cursor in the tag field on the search page and pressing 'search'), the list has some few entries for fiction and science fiction at the top. The difference seems to be that they were entered as +fiction and +science+fiction (having a + at the beginning makes them sort first).
Just thought I'd mention it.
Best wishes (and good luck with the server switch!)
sunny
And another detail about tags: does the total count get updated sporadically only? Not that I mind, just wondering. (Count on tag 'switzerland' has been at 111 for some days now. The numbers behind each user show up fine, but the sum is higher than 111). I had entered some via power edit and then one via normal edit to check: makes no difference).
sunny, LT addict ;-)
Hi, I just added tags to my (modest) collection... Around 400 tags for 80+ books.
I was wondering if there were any plans on having a nicer tagging interface -- showing a few common tags when you're adding them for instance.
And I really like that you allow multi-word tags :)
There's two ways of counting tags. I'm wondering what both counts are:
1) How many individual tags have been thought up?
2) How many actual tags are in use right now?
If that's not clear, "foo" is one tag for #1 but if it's used 10 times then it's ten for #2.
Unless it's a secret of course (-:
yay for 2.5 million tags!
i was just wondering if anyone else would like to see the global zeitgeist page reflect tag combinations? eg, it still says "...nonfiction 31,839...non-fiction 25,681..." shouldn't these be fused in the zeitgeist page just like authors and now editions are? this might just be my pet peeve, but i'd love to be able to see what the most popular tags *really* are.
nperrin
I want second the need for a nicer/easer tag interface. I think, like most users who first sign up, I am in the process of entering hundreds of previously read books. I usually enter major tags like science fiction, horror, mystery, etc. as I enter a new book. But what I think of as secondary tags like british, comedy, classic are harder to determine which to use so I plan to enter them later. What I would love to see is something which would recommend common tags already in use for a book and then let me select which I would like to use.
Another tag issue - there are some tags which have multiple spellings for instance, science fiction, scifi, sci-fi, sci/fi, sf. How are these handled? I started by using science fiction and then switched to sci-fi to save typing.
Thanks for a great job, keep up the good work.
mikeandtim
Tim, this is a really minor request... but I'd love it if you could give us the option of auto-alphabetizing our tags on individual books.
When I started out, I added tags in order of relevance or importance. But this system soon went askew due to some tag editing I did later on (which had to be done because of those tag limit issues). Right now, they're all over the place.
If I could view each entry's tags in alphabetical order, it would make it easier to identify which tags are yet to be added.
Of course, this should be optional, not forced on users... and if it has to be one or the other, then I'm happy with the status quo.
Minor sort observation: The $30,000 Bequest sorts to the beginning of an author's works; The £1,000,000 Bank Note to the end. (Both Mark Twain.) Any chance of getting £ up there at the beginning too?
LibraryThing may have 2.5 million tags, but Switzerland has 131!
Hmm, now it does, yes. Not easy to know beforehand which are the things to remark because they might matter in any way ;-)
Thanks for the reminder, btw. I set it up so tags would update on a regular schedule.
I use tags as broad descriptors for how and where I group my books, which is more like the Library of Congress (or Dewey) classification system than the subject headings. I think when I finish entering everything, I might go back and add more detail.
On Amazon: from the little I see on product pages, occasionally, and in my own habits, I'd say it's not going nearly so well. I tagged a handful of items around Christmas, but usually don't bother. There never seem to be many tags, and most don't look informative or useful, except in a very limited and single-user sense (i.e.: 'brenda's birthday'). Nor are they being applied in large numbers per user, per item.
While my tags here are still in a minor state of anarchy, I use a lot and care about adding them. - Though I've struggled with finding the right patterns for them. And adding useful tags can take time. With my own books, I care; with 'theirs,' I don't.
Who wants to do all that for a seller, however appreciated?? - Their books are, after all, already categorized. Though I find the results here richer. On LibraryThing, tagging is not only personal, but part of something cooperative and fascinating.
Actually, I really enjoy the few nicely 'messy' books I hit: they're fun to tag, and I love the ability to separate or combine various elements, and see disparate books that still cover some of the same territory. Etc... I could enthuse about tagging, LibraryThing, and your 'similarly tagged' suggestions at length, but I will spare you. :) Just consider it done.
- And congratulations on the 2.5 million tags! I am doing my minor part toward 2 million books, with slow but dedicated buying.
Last time I played with Amazon's tagging it still seemed not many people were tagging seriously but they had added a feature I really liked, which was to see recommendations based on tags. I played with it for hours!
I played with it for hours!
Yeah, but you are indefatigable!
Yeah, but you are indefatigable!
Only with fun things! With dull things I am very fatigable.
I have been entering my comics in the database and a number of other folks ahve been, too. I will note that any comics that I know have been complied into graphic novels(like Gaiman's Stardust) I have been entering as the collection, and noting in the comments that I have it in issues, but that's really been for convenience.
I believe librarything could get more visitors if the page title were not "frozen" as "Library thing Catalog your books". I think if it changed according to the book titles it could do better.
Post a Comment
<< Home