Flickr Patenting "Interestingness"

tjcrowder noted that Boing Boing is reporting that Flickr has filed for a patent on a system for determining "interestingness". From the patent application abstract: "Media objects, such as images or soundtracks, may be ranked according to a new class of metrics known as "interestingness." These rankings may be based at least in part on the quantity of user-entered metadata concerning the media object, the number of users who have assigned metadata to the media object, access patterns related to the media object, and/or a lapse of time related to the media object." So basically, nobody else can use tags to label files. Totally original thinking from the folks at flickr. *cough*

umm... no?

They're not patenting tagging, and there's no reason to think you can't use tags if the patent is accepted. The patent may be really stupid, but if we're going to get editorial comments, can they at least make sense?

Re:umm... no?

Agreed.

I really wish Slashdot would stop doing this. Taking a patent, making the most ludicrous assumptions about the scope, and then criticising these assumptions as ludicrous. It doesn't help. It undermines the anti-patent argument.

Read the blasted CLAIMS of the patent, Taco...

I really wish Slashdot would stop doing this. Taking a patent, making the most ludicrous assumptions about the scope, and then criticising these assumptions as ludicrous. It doesn't help. It undermines the anti-patent argument.

Indeed. Specifically, they seem to read the abstract of the patent--or even a third-party summary of the abstract--and fixate on one or two words from it, and say, "OH NOES THEIR TRYING TO PATENT TAGGING!!11!!!" or, "I have prior art for that! I saw someone with a tagging system back in '95!" when the actual claims for the patent (you know, the part that says what's actually patented) say something quite different, very specific, and not particularly worrying at all. Or even when they say something quite different, overly broad, and somewhat worrying...it still doesn't help, because you're arguing a completely different issue.

I am as distressed with the real broken state of patent law in this country as most others here...but the way in which it is treated is, as you say, very counterproductive. It would be very nice if there were some kind of standards for acceptance of such articles--and, though I'm generally not too critical of the editors, it would be much nicer if they would pay enough attention so as to not make completely worthless and off-base comments when they post them. Commenting is fine, just make it an informed, useful, and correct comment.

And no, before you ask, I'm not new here, I'm just annoyed.

Dan Aris

Commentary wrong ;)

So basically, nobody else can use tags to label files. Totally original thinking from the folks at flickr. *cough*

No, it just means you can't tag a file "interesting" ;)

Yahoooooo.

For those unfamiliar with Flickr, it's owned by Yahoo, who just [slashdot.org] loves [webmasterworld.com] them [searchenginewatch.com] some [typepad.com] patents. [resourceshelf.co.uk]

Not quite. . .

To the original poster of this article:

They're not trying to patent the use of tags. They're trying to patent a metric for measuring and quantifying meta tags.

The weird thing about this is that their metric doesn't actually measure the media itself, only the quantity of meta tags. I guess it works as a metric, but it's more a measurement of popularity, not "interestingness".

Somebody call Stephen Colbert

And make sure he patents 'truthiness'.

Deliciousness

Guess we'll have to start modding with +1 delicious now

Captain Obvious Strikes Again

Don't you just love how many of these companies go about patenting an idea (or similar idea) that is already used by the masses. This is such a joke, next thing you know (providing the patent is upheld) Yahoo will start suing folks like Amazon and our beloved Slashdot for patent infringement. Most people wouldn't even consider patenting something that is already, to some degree, "common knowledge". I thought the whole purpose of patenting was to protect intellectual property that has yet to be implemented or conceived. Hmmmm...I guess I'm just totally wrong in that assumption.

I think that's an oversimplified summary.

Tags are only a part of the interestingness. They want to patent a method for looking at patterns of popularity over time.

Part of their patent has claims on methodology, and part on a computer program designed to make use of that methodology (to cover the implementation requirement, I'm sure). And as I was reading just the initial page, I could imagine a pseudoequation forming in my head using the variables of time, popularity, content, etc.

I'm no patent attorney, but this sure sounds like trying to patent an algorithm. Not tags.
Now, whether they should do this is a good topic for debate -- but let's make sure we at least know what they're doing before debating it.

Striking back at Flickr

(for sufficiently low values of "striking back")...

Anyone who feels like sending a message to Flickr, join my new down with the interestingness patent [flickr.com] protest group.

(Not that anyone will listen, really, but...)

What an idiotic link!

Here is, in its uncut and undiluted entirety, the whole article that the summary links to:

Flickr files a patent for "interestingness"
Link [uspto.gov] to USPTO filing dated October 26.

That's it? WTF? Why not just directly link to the patent office and skip the ad-ridden Boing Boing link in the first place?

Finally!

At last we have the perfect /. article, since the vast majority of posters never RTFA anyway: an article that's completely summed up in the title of the /. post.

Think of it, guys. Not only don't you have to read the article, but you don't need the /. post, either! I'd call that real progress. Congrats, BoingBoing on a job well done!

Prior art

Isn't this http://www.iptc.org/pages/index.php [iptc.org] prior art?

Shouldn't slashdot be patenting this, or atleast..

Or at least mod points?

Interestingness is ... interesting

I'm not sure whether it's (or should be) patentable, but Flickr Intestingness is definitely a novel and, er, interesting concept. It's widely misunderstood even by hardcore Flickr-ites.

As far as I understand, an "Interestingness" score is derived from hits, referrers, tags, pool membership, comments and where comments come from, "favourite" tags and other things. The weighting is constantly being tweaked, and Interestingness changes over time because (for example, and hypothetically) a recent comment is more valuable than an old comment.

A number of great photographers get upset because they take high quality photographs which get lower Interestingness scores than pictures they perceive as having less merit. But Interesting is not about quality or merit. That's why it's not called "quality" or "merit". It's called "interestingness", meaning "cool stuff you might not have seen before".

That's why (again, for example and hypothetically) the tags "cat" or "baby" or "flower" are likely to have a negative impact on Interestingness. You can take the greatest baby picture ever, it's still not going to be interesting to most people, because Flickr is flooded with baby pictures.

In summary - it's cool, it's clever, it's more than just tagging, and it's novel.

I'd rather it wasn't patented but, hey, that's life.

Well, /. has no fear

Nothing "interestingness" here to see.

New Algorithm

if (average(imgAry.pixelval) == SKIN_TONE)
          return true;
else
          return false;

What's a patent about...

It is about the claims.

So when we check whether it makes sense, we check the claims.

1. An apparatus for determining an interestingness rank for at least one media object, comprising: logic for accepting at least one metadatum concerning the at least one media object from at least one user; and logic for ranking the at least one media object based at least in part on the quantity of user-entered metadata concerning the at least one media object.

Sure, my computer does not infringe as long as I don't have the software on it. And even then, would logic also be explained as software?

30. A method for determining an interestingness rank for at least one media object, the method comprising: accepting metadata concerning the at least one media object from at least one user; and ranking the at least one media object based at least in part on the quantity of user-entered metadata concerning the at least one media object.

In virtually any product review site, you can sort on user feedback. Anticipated; dependent claims do as far as I see no interesting embodiments.

54. A computer program product including computer-executable program code for determining an interestingness rank for at least one media object, the product comprising program code for: accepting at least one metadatum concerning the at least one media object from at least one user; and ranking the at least one media object based at least in part on the quantity of user-entered metadata concerning the at least one media object.

Same story as the method claim.

Apparently, Flickr thought it would be nice to spend a lot of money on a patent attorney just to have a patent application. Because I really wonder whether this will turn in a patent.

But if they patent it...

...are they not supposed to describe in detail how it works? Part of how "Interestingness" works is that no one knows exactly how it is determined, so that it is difficult for someone to exploit the system. Also, the exact combination of inputs changes regularly to keep things... um, interesting. Is it really enough for them to loosely describe a system where they automatically determine a rating from a collection of inputs to get a patent? And if so, should it be? It seems just about anything can get a patent now, which is really a shame. I think the patent system could work just fine if it were managed properly, but in its current state, it certainly seems like it is hurting more than helping innovation... -- plasmoidia

Somebody should patent patenting

Somebody should patent the action of patenting things. Then everybody that gets a patent should be forced to pay that person royalties.

Let Me Be the First...

to say, "A patent application on 'interestingness?' Boooooriing."

IANAL

But doesn't something like Google Trends already implements at least part of the patent? If that's the case, shouldn't prior art easily defeat that patent, since it's invalid anyways?

Thermonuclear patent war

Software patents are like nuclear warheads. Companies amass a whole bunch of them so that if somebody starts the war, they will have retailatory salvo. If Microsoft sues Yahoo for patent infringement, they counter-sue. It's about increasing the risk to the potential claimant--mutually assured destruction.

There's another application

There is a co-filed Companion Application [uspto.gov] which has the following as its broadest claim:

1. An apparatus for associating metadata with at least one image, the apparatus comprising: logic for providing the at least one image for display to at least one user; and logic for accepting input from a plurality of users concerning the at least one image, wherein the input comprises at least two different types of metadata.

The metadata types referred to are tags, comments, descriptions, favorites and annotations. So, this claim would cover presenting one or more images to a user and then accepting at least two of the types, such as tagging and commenting. Claims 1 - 27 are directed various details of this.

Claims 28 - 142 are where the applicants want to really make money from this concept:

28. An apparatus for providing information for use by an ad server, the ad server for associating an advertisement with a media object, the apparatus comprising: logic for providing the media object for presentation to at least one user; logic for accepting metadata from a plurality of users concerning the media object; and statistics logic for determining a metric based at least in part upon the frequency with which at least one metadatum is assigned to the media object, wherein the at least one metadatum is available to the ad server.

Basically they are allowing an adserver to select an ad to show with the media object by sending at least one metadatum to the adserver based on their crunched metric.

Interestingness is to Flickr

what the Google search algorithm is to Google. In case there's no die-hard flickr-rers in here, Interestingness is really sorting out chaff from wheat, as regards to photography. It's not just tagging stuff, folks -- it's separating good photographs, funny photos, *cough* interesting photos, artful, emotional photos from the rest of the garbage sent into flickr.

Try this experiment: Search Flickr for a thing, concept or emotion. Sort by "Most Recent." Bask in the great number of bad snapshots. Now take the same search, and sort by "Interesting". Whoa -- art. And if not art, at the very least, "interesting".

No one other than the writers know how it works. Flickr don't say how it works. A few points, based on observation:

o Interesting has nothing to do with art, really -- but artful shots are bound to be more interesting than typical bad snaps.

o In scale of weight: At the top there's Favorites, then Comments, and finally, Views.

o I suspect the algorithm also somehow interprets the photo on its own, with no human help. Probably looking for photographic concepts such as tonality, composition (rule of thirds, etc.) I feel this because I've uploaded images which I would say should rank very low, yet almost immediately they're ranked higher than I expected -- with no views, no comments. This suggests to me Flickr's algorithm is actually looking at the pix. If this is true, then it's remarkable.

I say they have every right in the world to patent it. It makes Flickr different than the other 'sharing' sites.

Sour grapes...

Slshdot is kicking itself for not patenting its moderation system?

Moo

In related news, Slashdot will be removing "+1 Interesting" in compliance with the patent. Moderators are reccomended to use "underratred" instead. Officials under condition of anonymity report that there won't be much of a difference anyway because noone uses Interesting anymore duie to meta-moderation.

flickr discussion ...

You can find a very good discussion on the flickr group, utata:

http://www.flickr.com/groups/utata/discuss/7215759 4355879087/ [flickr.com]

I submitted this story 1 week ago. Old news.

No problem

I'll just have to use interestingosity instead. Or maybe interestingological. Interestingologicalosityness?

Is it the same as truthiness?

I think patenting interestingness infringes on a similar patent someone already holds on truthiness.

One patent to rule us all...

Flickr will rule us all!

Re:Thermonuclear patent war

Patents should be approved and managed like trademarks; in order to be valid and defendable one must show a serious attempt to keep it in active use, else it is expired/invalidated.

Re:Can we just ...

If you ban discussions of patents here, you'd just get the same Slashbots screaming "Infringment of First Amendment rights!" and "Censorship!" without thinking that, hey, it's a privately-owned website.

Unless an asteroid hits.

How badly do we want to handicap American ingenuity - do we really want to eventually hand the industry to Europe and Asia (places without barratry- I mean, software patents)?

It reminds me a little of another industry that was one the 'crown jewel' of the U.S. economy: the automobile industry. The unions and management never seemed to stop and take a look around, even as most of their jobs and business moved to non-union factories in Asia (which, in addition to being cheaper, produced an arguably superior product). Even today, they're still bickering over the bloated, living-dead corpse of GM.

So to answer your question, I fully expect that the idiocy over software patents will end when the last American software programmer hangs himself with the entrails of the last American intellectual-property attorney, and not a moment sooner.

Re:Software patents absurd

Have you seen the sorts of patents the USPTO allows? How about patenting right click? Yeah, they have allowed that. Also the PTO has a LOUSY record of recognizing prior art with software with thousands of "patents" ending up invalidated by obvious prior art. Microsoft has a patent on human skin, because it conducts electricity and might, just maybe, somehow, someday be used to transmit computer signals. There's no specificity in that patent at all. The PTO is manned by lobotomized monkeys who wouldn't know an original idea from the 5 billionth printing of a Bazooka Joe comic.

Re:What other tags?

Too late, I think Stephen Colbert has already filed for that one.

...eyes... hurt...

<BR>! <BR>, damn you!

For the love of God, man... please... <BR>...


<whimper>