Natural Language Processing for State Security

Roland Piquepaille writes "Obviously, computers can't have an opinion. What computers are very good at, though, is scanning through text to deduct human opinions from factual information. This branch of natural-language processing (NLP) is called 'information extraction' and is used for sorting facts and opinions for Homeland Security. Right now, a consortium of three universities is for the U.S. Department of Homeland Security (DHS) which doesn't have enough in-house expertise in NLP. Read more for additional references and a diagram showing how information extraction is used."

Comments

tinfoil hat... or is it?

[macadamia_harold]

What comptuers are very good at, though, is scanning through text to deduct human opinions from factual information. This branch of natural-language processing (NLP) is called 'information extraction' and is used for sorting facts and opinions for Homeland Security.

Yeah, because we need AT&T giving wide-scale, undocumented wiretaps to the NSA, who use voice recognition to generate transcripts of everyone's phone calls, and then DHS can run NLP on those transcripts to compile a list of "persons of interest", who are then automatically added to the TSA no-fly lists.

Yeah, I can envision the future, and the future sucks.

Re:tinfoil hat... or is it?

[rtb61]

Sorting facts from opinion by use of language, how amazingly pointless and stupid. Now lets see if the program can sort BS facts from real facts. This just seems like another scheme cooked up by incompetant political appointees, who don't have any idea about what they are being paid to do. Their only hope of retaining their postition, so they can continue their real function of politcal party support for the current adminsitration, is to try to get that magic box to do their job for them.

You want to know how incapable they are, look at the extent and perversion of punishment inflicted upon innocent people at GITMO prior to any semblance of justice being done i.e. proving the truth of the claims and opinions of a whole swag of political appointees in court of law (you are innocent until proven guilty in a court of law, precisely because past generations learned the lesson that you can not trust the enforcers of the law unless they are held under constant public attention and review).

Exactly how many successful prosecutions have there been after all those years of operation of a facility that is clearly a perversion of jusctice. They are more likely to get successful prosecutions, against those who created and operated that facility, rather than the inmates (it is not about the accussed terrorists, it is about ensuring that the government adheres to the principles of justice, so that future generations do not suffer the perversions of justice that past generations suffered).

Moo

[Chacham]

What comptuers are very good at, though,

.... is spell-checking.....

....something, apparently, the editors are not good at....

Re:Moo

[ceoyoyo]

Maybe Roland had a stroke over the weekend. Sure he's self serving, but at least he's usually literate. That sentence about the universities didn't even make sense!

Sigh.

[Renraku]

The slippery slope to being automatically flagged as someone to watch out for. No human control in the process, but one day when you go to apply for a loan or get your drivers' licence renewed, you might get a surprise.

Re:Sigh.

[BiggerIsBetter]

I'd rather have a computer flagging me than a human who may judge me by the color of my skin

If they can flag based on what you said, I'm sure they can flag you based on the skin tone in the photo on your drivers license or passport too. Or by your just family history or name. Or where you live. Or where your parents live.

Anyways, odds are the computer won't be doing the flagging per se, it'll just be following the parameters and policies entered by those humans controlling it. I'm not sure they'd trust "national security" to a self-learning neural net without some sort of bias in it.

Number 891224

[bky1701]

Number 891224 has expressed a dislike of Emperor Bush, incident reported to FBI and Homeland Security.

Re:Number 891224

[ikkonoishi]

Number 979071 has expressed an interest in Emperors, incident reported to George Lucas

Alias-i's ThreatTracker

[otisg]

There is a great little company in Brooklyn, NY called Alias-i. Some years ago they built this interesting "tool" called....guess....ThreatTracker. Information Extraction, Named Entity Recognition and other interesting stuff, if you are into this.
No, I don't work for them, but their LingPipe toolkit has some cooooool stuff.

really?

[agendi]

"Obviously, computers can't have an opinion. What comptuers are very good at, though, is scanning through text to deduct human opinions from factual information."

I would say that comptuers (sic) aren't very good at deducting human opinions yet. They _may_ become better. Are humans good at deducting other humans opinion yet?

A really difficult problem

[MarkWatson]

I have, in agregate, spent about 3 1/2 years in the last 20 years working on using NLP for semantic information extraction.

Possible? Yes, given very narrow domains of discourse and lots of work.

Re:A really difficult problem

[Walt Dismal]

I agree with the 'lots of work' part, but believe it is possible to achieve good results on wider domains outside of toy worlds. One key - from my own research - is to use (massive) databases of culture-related knowledge (belief systems) to build alternative viewpoints from which to massively parallel analyze the input. Each analysis agent has its own viewpoint or frame, driven by a very large database of world knowledge that is culture-specific. By culture I mean not just nationality but specific domains of belief systems. For example, American+middle class+scientist+age range 40-50, or white male Protestant businessman :) etc. Derrida was kind of right in a way; you have to bring specific personal knowledge to interpreting something, and no two people come to anything in exactly the same way. But two people with similar cultural bases will see similarly, all other things being equal.

The system has to handle complex contexts and multiple varying worldframes. It has to superimpose multiple viewpoints - alternate personnas - in interpreting the source. Also useful is applying certain theories of story to modeling the world.

Yes, it's non-trivial, but achievable. I can see a day coming where a Google-like entity can, by modeling you then acting as a 'cloned' agent, apply such personnas in your service and find not just data but meaning, for your benefit.

Re:A really difficult problem

[constantnormal]

It's the "narrow domains" that is the crux of the problem.

When used successfully over said "narrow domains", the human tendency (especially that set of humanity which makes the high-level choices for groups and organizations) will be to expand the domain in hopes of applying it to ever greater numbers of items.

Of course, as the search domain is expanded, the effectiveness of the results decline, with no warning to the clueless idiots driving the search. False positives eventually exceed true positives by greater and greater margins.

In the end, the strategy collapses, as a great many victims are shown to be wrongly targeted -- but until that point, the system does a LOT more harm than good.

Thank Goodness our leaders are such wise and contemplative souls that they would never, ever misuse such a tool.

A boon to research

[JanneM]

It's clear "national security" has become what "the internet" or "the cold war" were in their prime: an all-purpose catchphrase to get funding for any research whatsoever, no matter how tenuously connected.

Look at the two project proposals below and imagine which one will have an easier time getting funding:

"An epistemological metaanalysis of object-subject interrelations and conflict avoidance in Beowulf"

or

"An epistemological metaanalysis of object-subject interrelations and conflict avoidance in Beowulf to better understand threats to NATIONAL SECURITY"

Re:A boon to research

[NitsujTPU]

With all due respect, that is inaccurate.

DARPA, Defense Advanced Research Projects Agency, is a gigantic agency that funds a large proportion of academic research. The political hot button of child pornography, on the other hand, has no large funding source to offer universities. That's why so many academic projects have ties to defense.

Also, yes, usually research is, "do whatever you were going to do, but tie it to defense somehow." That's the way it goes, you need the cash. However, usually you can tie fundamental research to defense in some way. One of the PhD students who was at Cornell while I was there used movie reviews for related research... however, the simple mark "positive" or "negative" is certainly enough to help the DoD filter Internet documents if they chose to do so, so there is a tie to defense. The technology had a reason for existing without the DoD, but funding might have been another story. The same goes for cars that drive themselves, humanoid robots, and distributed computing (though distributed computing has nice corporate interests through companies like Amazon and Google that have to maintain mega-networks for their operations).

Two Roland junk submissions in two days

[Anonymous Coward]

Wow, thanks for another waste of time. And you people stop linking to his blog in comments, he exists for nothing but ad clicks.

Re:Two Roland junk submissions in two days

[Anonymous Coward]

Not to mention, he linked to the almost EXACT same blogs he did last night in his Hydrogen junk article, tisk tisk Roland. You can mod us offtopic all you want man, just checking the last article proves he is scamming Slashdot (and it's users) for ad clicks for these blogs and his own.

Let us filter this guy please. Seriously, I will stop subscribing and so will my usergroup if we can't filter out his faux science crap. It's getting near the end of the month Slashdot, do you, Roland ,and his ad carriers have some deadline to beat?

In case anyone was wondering why some of us keep bitching, it's because we have NO other way to get our point across short of no more subscribing here! Some of us actually like to discuss REAL science, not this crap Roland keeps getting on the front page here. We can't filter it, so we get raped by his ads and ad partners, OR we ignore science here on Slashdot and get real bored real fast ;/

Man...

[GnomeChompsky]

There goes a promising career path. I know any technology can be used for good or for evil, but in today's political climate, it seems especially irresponsible to be aiding and abetting what may wind up becoming the pretext for torture of some 16 year old blogger.

Now, if you'll excuse me, I have to prepare myself for my upcoming extraordinary rendition....

Sounds like GALE

[Dr. Eggman]

Sounds kind of like DARPA's Information Processing Technology Office's GALE Program:

" The goal of the GALE (Global Autonomous Language Exploitation) program is to develop and apply computer software technologies to absorb, analyze and interpret huge volumes of speech and text in multiple languages, eliminating the need for linguists and analysts and automatically providing relevant, distilled actionable information to military command and personnel in a timely fashion. Automatic processing "engines" will convert and distill the data, delivering pertinent, consolidated information in easy-to-understand forms to military personnel and monolingual English-speaking analysts in response to direct or implicit requests."

Yes, computers are great at spell checking

[Colin Smith]

That doesn't stop the really determined idiot though. Oh no.

        I have a spelling checker,
        It came with my PC.
        It plane lee marks four my revue
        Miss steaks aye can knot sea.

        Eye ran this poem threw it,
        Your sure reel glad two no.
        Its vary polished in it's weigh.
        My checker tolled me sew.

        A checker is a bless sing,
        It freeze yew lodes of thyme.
        It helps me right awl stiles two reed,
        And aides me when eye rime.

        Each frays come posed up on my screen
        Eye trussed too bee a joule.
        The checker pours o'er every word
        To cheque sum spelling rule.

        Bee fore a veiling checker's
        Hour spelling mite decline,
        And if we're lacks oar have a laps,
        We wood bee maid too wine.

        Butt now bee cause my spelling
        Is checked with such grate flare,
        Their are know fault's with in my cite,
        Of nun eye am a wear.

        Now spelling does knot phase me,
        It does knot bring a tier.
        My pay purrs awl due glad den
        With wrapped word's fare as hear.

        To rite with care is quite a feet
        Of witch won should bee proud,
        And wee mussed dew the best wee can,
        Sew flaw's are knot aloud.

        Sow ewe can sea why aye dew prays
        Such soft wear four pea seas,
        And why eye brake in two averse
        Buy righting want too pleas.

  -- "Candidate for a Pullet Surprise"
By Jerrold H. Zar, Northern Illinois University
Journal of Irreproducible Results 39, 1 (Jan.-Feb. 1994): 13

abuse?

[mr100percent]

Why do I immediately assume this will be abused?

DHS officer: Mr. 100%, I'm afraid we'll have to take you into custody. Our information extraction search on your blog concluded you are anti-American.
Me: From my blog? Is this about my criticism of the Iraq war?
DHS officer: Our results are classified, but please accompany us to GTMO for further "information extraction" to confirm the results of our investigation...

Ok, I know I'm taking a very cynical view here and that's pretty full of FUD, but why else does State security need this? Is this for them to monitor every chat room and blog?

Aha!

[suv4x4]

Obviously, computers can't have an opinion.

Welcome the new opinion-based CAPTCHA-s!

Can do or will do?

[Dan East]

What comptuers are very good at, though, is scanning through text to deduct human opinions from factual information.

Funny, because neither of the articles state that. In fact, they don't even say that software can do that at all yet: A new research program ... aims to teach computers to scan through text and sort opinion from fact. Or, We're interested in seeing how we would extract information about opinions.

So yeah, it would be nice if they could sort opinions from facts. Why they're at it, why don't they just recognize lies from truth too, because wouldn't that be doing the exact same thing? Then we can just run statements made by people suspected of committing a crime through the software, which can then sort out all the facts from the opinions, and we'll no longer need judges, juries or attorneys.

Roland, next time save yourself some time and just make the whole freaking thing up from scratch.

Dan East

one thing

[shack420]

another thing Rolands computer is not very good at is spell checking his posts!

screw national security

[argoff]

Screw national security, how about search, how about for business and commerce, how about for for culturial exchange and global interaction. The chances of me getting attacked by a terrorist are less than getting hit by lightning, the chances with dealing with foriegn cultures, foriegn business and commerce are rapidly approaching 100%. There are 4 billion people out there who have the potential to mutually benifit from clean communication. Please don't patrinoze me, I'm not too worried about getting nailed by terrorists, but am very bothered by the possibility of having my individual liberties nickeled and dimed to death.

Re:Ooooh another funding scramble!

[Mac Degger]

Especially since the system, whilst it will have some quite interesting applications and the research will yield interesting results, can't work. A computer cannot distinguish between a fact and a lie told as fact...garbage in, and all that.

Let me rephrase that with an example:

'I am ten years old' and 'I am twenty years old'. Which is fact, which is lie? Better yet: 'we believe Iraq has WMD' versus 'we beleive Iraq has no WMD'. No matter what algorythms or heuristics you throw at this, all a computer at most can tell you is 'sometimes when used in conjunction with this phrase, the statement is false'...but that helps you IN NO WAY, because it means the statement can also be true...the indicator means nothing...you get as many false positives as false negatives...hell, even a ratio would be meaningless in intelligence gathering.

Bushed

[Anne Thwacks]

You mean it was not the computers that voted for George W Bush? Then who the hell did?

Information extraction

[pk075842]

Information extraction (IE) is a type of information retrieval whose goal is to automatically extract structured or semistructured information from unstructured machine-readable documents. A type of concept extraction that automatically recognizes significant vocabulary items in text documents, such as, names, terms, and expressions.