Prev: recent formal proof re: seL4 microkernel
Next: Question on queuing theory
From: Mark Murray on 28 Nov 2009 05:08
JSH wrote:
> Consider 2+2 = 4.
>
> OPINION can be that 2+2 = 5.

In your case, that is tantamount to fact.

> Human beings can think the damndest things.

Indeed. You think your SWJPAM paper is correct.

> Google search results would tend toward 2+2=4 is the hypothesis, even
> if people generally believe that 2+2=5.

No. If there were enough people saying 2+2=5, 2+2=5t would get a high
ranking, irrespective of its correctness. Google doesn't judge, Google
counts.

Look at the high ranking some conspiracy theories get.

Google simply "twin towers" and see how far from the top the first bit
of conspiracy theorising comes.

> IN context, here people like you on Usenet insult me, but my research
> nonetheless beats top people in the field in areas of their expertise,
> like: solving quadratic Diophantine equations

Alpern beats you hands down. His process is complete and practical.

> So DESPITE human opinion, Google is indicating that the group effect
> is to find the truth, or you may disagree and say it's a worthless
> indicator, which challenges the worth of Google searches themselves.

/non sequitur/

> Google can be considered to be an objective, disinterested third
> party.
>
> Academics challenged by my research or people challenged by it for any
> number of reasons can be considered interested first or second
> parties.
>
> IN a word, you may be biased.
>
> Prejudice is as old as humanity.

So is selection bias, so is narrow-mindedness.

Just because someone else /may/ be wrong, doesn't mean you aren't.

M
From: Joshua Cranmer on 28 Nov 2009 17:45
On 11/28/2009 5:08 AM, Mark Murray wrote:
> Look at the high ranking some conspiracy theories get.
>
> Google simply "twin towers" and see how far from the top the first bit
> of conspiracy theorising comes.

"twin towers fall" produces two results of the official theory and five
of the conspiracy theories. "collapse of world trade center" produces
three results of the official theory and three of the conspiracy theory.
Both results were counted as the number of non-indented result pages in
the first ten results. Note that the terminology produces a difference:
the latter result turned up the NIST report, an article (though not a
scientific paper) from a materials science trade publication, and a
Wikipedia article as the pro-official searches, whereas the former
result turned up a Wikipedia article and an About article.

Bringing this to a point: using the more proper terms will turn up more
scientific results [1], a metareview of which would show you that the
conspiracy theories are probably wrong. Colloquial terms will return
less scientific results, which will defer the results to be biased more
by the weight of public opinion.

In short: if you want to find the truth, eschew colloquialism in your
searching.

[1] One can argue whether or not Wikipedia can be considered scientific
in this context. I've never verified sources, but the Wikipedia entries
on highly controversial issues (like the WTC collapse) contain extensive
source sections and have highly approachable text. I list Wikipedia here
as pro-official theory, since it covers the official theory more
extensively, the link goes to the official theory, and the conspiracy
theory articles read as if somewhat dismissing them as crackpot.

--
Beware of bugs in the above code; I have only proved it correct, not
tried it. -- Donald E. Knuth
From: JSH on 28 Nov 2009 18:19
On Nov 28, 2:45 pm, Joshua Cranmer <Pidgeo...(a)verizon.invalid> wrote:
> On 11/28/2009 5:08 AM, Mark Murray wrote:
>
> > Look at the high ranking some conspiracy theories get.
>
> > Google simply "twin towers" and see how far from the top the first bit
> > of conspiracy theorising comes.
>
> "twin towers fall" produces two results of the official theory and five
> of the conspiracy theories. "collapse of world trade center" produces

And still you don't have what I'd consider a natural search.

A better search is: twin towers collapse

The top 10 results are enlightening. Wikipedia leads and I'd say that
NONE are conspiracy theories.

Cherry-picking ways of doing the searches to make an unnatural way is
not hard to do!!!

That doesn't change the authority level for results from how people
ACTUALLY search, but it is of interest to watch posters engage in the
exercise.

> three results of the official theory and three of the conspiracy theory.
> Both results were counted as the number of non-indented result pages in
> the first ten results. Note that the terminology produces a difference:
> the latter result turned up the NIST report, an article (though not a
> scientific paper) from a materials science trade publication, and a
> Wikipedia article as the pro-official searches, whereas the former
> result turned up a Wikipedia article and an About article.
>
> Bringing this to a point: using the more proper terms will turn up more
> scientific results [1], a metareview of which would show you that the

I'd say doing a search in a way that people who actually want
information might do it, is what works.

Posters arguing with me before would come up with wacky searches,
unlike ANYTHING that a person trying to actually get information would
use, and then chortle claiming victory when they got bogus search
results.

But that's just a form of GI-GO: Garbage In-Garbage Out

> conspiracy theories are probably wrong. Colloquial terms will return
> less scientific results, which will defer the results to be biased more
> by the weight of public opinion.
>
> In short: if you want to find the truth, eschew colloquialism in your
> searching.

Like with the following search: define mathematical proof

It is succinct, to the point, and is a request from the user. I use
such requests myself all the time to get definitions.

In this case it gives my definition of mathematical proof--my personal
definition.

To my knowledge I am the only person in the world who can give you
such a search with a definition of such importance!!!

Making me 1 in approximately 6.8 billion people on the planet who can
do that.

> [1] One can argue whether or not Wikipedia can be considered scientific
> in this context. I've never verified sources, but the Wikipedia entries

I beat Wikipedia with that search.

They were #1 for quite some time until I gained on them and ultimately
beat them out, which dropped the Wikipedia to #3 for a while until it
recovered to #2, after some months.

I've talked about this issue for a while and as the Wikipedia
continues to update my hope is that changes to that page might have it
beat me once again.

I emphasize that I beat the Wikipedia, and in fact my competition is
often the Wikipedia. For me it is often just a rival where it wins
sometimes and I win sometimes.

But Microsoft can't beat me either with Class Viewer.

They have a Class Viewer for .NET, while mine is for Java.

My open source program roared to the top years ago, and has not been
beaten since.

It continually beats out any number of other applications with the
same name, including Microsoft's.


> on highly controversial issues (like the WTC collapse) contain extensive
> source sections and have highly approachable text. I list Wikipedia here
> as pro-official theory, since it covers the official theory more
> extensively, the link goes to the official theory, and the conspiracy
> theory articles read as if somewhat dismissing them as crackpot.

Google search results continue to improve in my opinion.

I think readers would be hard pressed to give a Google search that
puts up a crackpot theory highly in any area where most people do not
subscribe to that theory, with a sensible search!

I use searches that make sense, like: solving binary quadratic
Diophantine equations

Readers would be hard-pressed to find research from ANYONE in the
world that beats me, IF you are actually searching on how to solve
binary quadratic Diophantine equations!

My guess is that academics are simply being crushed out because what
they have is worse.

Google is acting as an objective third party, which is revealing I
suggest serious flaws in what academics will teach as the best in the
world.

For instance, you cannot find a better way to generally handle binary
quadratic Diophantine equations than what I give.

That is just a truth. Challenges to my assertion are not to call me
names, or to claim the search is bogus, but to GIVE a better method
for handling them!!!

But you see, the trouble for Usenet posters in denial is that there is
none.


James Harris
From: Joshua Cranmer on 28 Nov 2009 18:54
On 11/28/2009 6:19 PM, JSH wrote:
> A better search is: twin towers collapse
>
> The top 10 results are enlightening. Wikipedia leads and I'd say that
> NONE are conspiracy theories.

No, the "wtc7.net" one is most definitely a conspiracy theory site. The
banner practically screams it ("misinformation"). Just reading some of
the stuff on that site makes you want to pull your hair out (the
cellular system in the U.S. is not designed to handle a cell phone
moving at 10K feet at several hundred miles per hour...).

> I'd say doing a search in a way that people who actually want
> information might do it, is what works.

You completely miss the point, again. The thesis of our argument is that
the search results of Google has no bearing on the validity of the
pages. Indeed, what has just been demonstrated is that, in some cases,
the "natural search terms" produce invalid results.

> Making me 1 in approximately 6.8 billion people on the planet who can
> do that.

Quite the egotist. Proof by failure to find a counter-example is not,
after all, a proof. It's just a mixture of laziness and, in some cases,
egotism or perhaps elitism.

> My open source program roared to the top years ago, and has not been
> beaten since.

Try jshydra. Wait, there's no competition for that name. How about the
other projects I work on:
* Thunderbird
* Dehydra
* DXR

Wow, all of them in the top 10 search results. Two of them beat out
other usage names. Weird coincidence, isn't it?

> My guess is that academics are simply being crushed out because what
> they have is worse.

How about because they don't use the search terms that you do? Many
fields have more technical terms instead of more common place
terminology. Googling for "CAPTCHA" is more useful than looking for
"image spam verification" (trying to find the name of CAPTCHA was a
pain...). I'm more likely to find chemical information on dynamite by
looking for trinitrotoluene instead of TNT or dynamite.

> That is just a truth. Challenges to my assertion are not to call me
> names, or to claim the search is bogus, but to GIVE a better method
> for handling them!!!

You're trying to argue that searches are validation of your methods, and
then claiming that the only refutation we can give is to say that your
methods are wrong. I'll stick to disproving your line of argument rather
than disproving your overarching hypothesis. That shifts the burden of
proof to you to show that your method is correct... by mathematical
logic, not proof by authority.

But I digress.

--
Beware of bugs in the above code; I have only proved it correct, not
tried it. -- Donald E. Knuth
From: JSH on 28 Nov 2009 22:18
On Nov 28, 3:54 pm, Joshua Cranmer <Pidgeo...(a)verizon.invalid> wrote:
> On 11/28/2009 6:19 PM, JSH wrote:
>
> > A better search is: twin towers collapse
>
> > The top 10 results are enlightening.  Wikipedia leads and I'd say that
> > NONE are conspiracy theories.
>
> No, the "wtc7.net" one is most definitely a conspiracy theory site. The

I didn't check the ENTIRE SITE, just the page that comes up highly
which is to a lot of video.

> banner practically screams it ("misinformation"). Just reading some of
> the stuff on that site makes you want to pull your hair out (the
> cellular system in the U.S. is not designed to handle a cell phone
> moving at 10K feet at several hundred miles per hour...).

And I didn't see any of that when I checked which I just did after
reading your comment.

That's how bias sneaks in--even a site that has OTHER PAGES that
scream misinformation to you could have a SINGLE PAGE that lots of
people find useful, and checking that page I could see why it might be
considered useful.

> > I'd say doing a search in a way that people who actually want
> > information might do it, is what works.
>
> You completely miss the point, again. The thesis of our argument is that
> the search results of Google has no bearing on the validity of the
> pages. Indeed, what has just been demonstrated is that, in some cases,
> the "natural search terms" produce invalid results.

Yet the #1 page is to a Wikipedia article!!!

Or do you claim that the Wikipedia article is in error?

Your primary point now is that a SINGLE page because it was on a site
that also contained pages you feel were wrong must itself be invalid,
when it was a page linking to videos from news organizations, and came
up #3, as it didn't even come up #1. The Wikipedia did.

> > Making me 1 in approximately 6.8 billion people on the planet who can
> > do that.
>
> Quite the egotist. Proof by failure to find a counter-example is not,
> after all, a proof. It's just a mixture of laziness and, in some cases,
> egotism or perhaps elitism.

Point taken. One might argue that there are any number of important
definitions out there, like: define quantum mechanics

And maybe with one of those some individual person's definition trumps
what is in the dictionaries.

In my case I have defined mathematical proof.

> > My open source program roared to the top years ago, and has not been
> > beaten since.
>
> Try jshydra. Wait, there's no competition for that name. How about the
> other projects I work on:
> * Thunderbird
> * Dehydra
> * DXR
>
> Wow, all of them in the top 10 search results. Two of them beat out
> other usage names. Weird coincidence, isn't it?

Like what? From whom?

I didn't mention DMESE which beats out a stock symbol as I was
emphasizing areas of competition.

Like with Class Viewer I beat Microsoft. THE Microsoft. The people
who make Windows. Heard of them?

Do you have any comparable competition?

With my definition of mathematical proof, I beat the Wikipedia. And
it's a hard fought battle as well!

When I finally topped them, they fell two spots to #3 before managing
to claw their way back to #2, behind my definition.

Do you have competition on the level of the Wikipedia?

> > My guess is that academics are simply being crushed out because what
> > they have is worse.
>
> How about because they don't use the search terms that you do? Many
> fields have more technical terms instead of more common place
> terminology. Googling for "CAPTCHA" is more useful than looking for
> "image spam verification" (trying to find the name of CAPTCHA was a
> pain...). I'm more likely to find chemical information on dynamite by
> looking for trinitrotoluene instead of TNT or dynamite.

Or the definition of mathematical proof?

Oh yeah, I also take: definition of mathematical proof

> > That is just a truth.  Challenges to my assertion are not to call me
> > names, or to claim the search is bogus, but to GIVE a better method
> > for handling them!!!
>
> You're trying to argue that searches are validation of your methods, and

Nope. I'm not.

> then claiming that the only refutation we can give is to say that your
> methods are wrong. I'll stick to disproving your line of argument rather
> than disproving your overarching hypothesis. That shifts the burden of
> proof to you to show that your method is correct... by mathematical
> logic, not proof by authority.

Burden of proof? Why do I need that?

Do you actually believe I'm trying to convince you my research is
correct?

Why would I bother? Who are you?

> But I digress.

The issue here is not what you personally believe. I don't care if
you believe in the Tooth Fairy.

The question is, does Google as an objective third party represent an
"authority" which can trounce, say, the academic journal process?

Just what is "formal peer review" in the Internet web search age?

If people are directed to me as #1, versus academic pages, which
academics are claiming are the best available, is that an error by
Google? Or an error by academics to update to the best research
available in the world?

I think objective analysis and skeptical consideration would deem it
LESS likely that search engines like Google would carefully select out
bogus research versus mainstream academic research, versus academic
researchers dragging their feet to acknowledge the best new
information available.

So my position is that Google gives you the cutting edge of research.

Now I may be biased towards my own ideas, but I think I gave a rather
nice, succinct and useful definition of mathematical proof, especially
when compared with mainstream academic ones.

And my Class Viewer has the best competition available in the world,
and at least by its name, it trounces Microsoft's own product of the
same name, which is again for .NET, while mine is for Java.

Do you think Microsoft takes that lying down? That to them it's just
a minor funny thing? No big deal?

Have you ever competed with Microsoft?

I LOVE these discussions as the irrationality is mind-boggling.

Unless you see the irrationality as having a purpose!

Denial.


James Harris
 |  Next  |  Last
Pages: 1 2 3 4 5 6
Prev: recent formal proof re: seL4 microkernel
Next: Question on queuing theory