Science Algorithm for Learning Languages

[NiteFox]

Vaguely-interesting tech story.

http://www.scienceblog.com/cms/node/8802

Cornell University and Tel Aviv University researchers have developed a method for enabling a computer program to scan text in any of a number of languages, including English and Chinese, and autonomously and without previous information infer the underlying rules of grammar. The rules can then be used to generate new and meaningful sentences. The method also works for such data as sheet music or protein sequences.

[Guest]

It sounds like imitating rather than learning.

It may be useful though and also could be a helpful short cut for loops learning to "talk".

[Occhidiangela]

Vaguely-interesting tech story.

http://www.scienceblog.com/cms/node/8802
[right][snapback]87946[/snapback][/right]

This sounds like the kind of stuff Naom Chomsky used to work on, before he became a pundit.

Occhi

[Lady Vashj]

Protein sequences have rules of grammar?

[--Pete]

Hi,

Protein sequences have rules of grammar?
[right][snapback]87973[/snapback][/right]

Yes. For a start consider 'generative grammar' which is "a description in the form of an ordered set of rules for producing the grammatical sentences of a language" according to M-W. In the case you bring up, the amino acids are the 'words'. A sequence of amino acids is controlled by the rules of electrodynamics (the 'grammar' of chemistry). All sequences of amino acids formed by those rules are 'grammatically' correct. Just as in natural language, not all the proteins so derived are syntactically correct (e.g., "The horse designed the sky." is a grammatically correct sentence, having its nominative, its predicative, and its direct object. However, in the hands of anyone not so skilled as Lewis Carrol, it is not a well formed sentence syntactically). As custom is the arbiter of syntax in natural language, so is mRNA the arbiter of valid structures in proteins -- or was till the lab rats went to work. ;)

So, in the final analysis, all we are is the rules of electrostatics applied to a handful of elements.

"Of course you are, my bright little star . . ."

"I'm more than that. At least, I think I must be."

(Name that tune :) )

--Pete

[Drasca]

or was till the lab rats went to work
[right][snapback]87977[/snapback][/right]

"Pinky and the Brain have taken over the world" -Anonymous, circa yesterday morning

[Guest]

I dont think thats what they were alluding to.

I think they meant that the algorithm was able to create amino-acid chains that formed functional regions similar to those found on natural protiens.

They are saying that a grammatically correct sequence would be one that forms at least a quasi-functional region.
To continue the analogy to written language a logically correct sentence(which this algorithm doesnt create) would be one that formed a region that could actually be used for a preconcieved function in a protien.

[Bun-Bun]

"Of course you are, my bright little star . . ."

"I'm more than that. At least, I think I must be."

(Name that tune :) )

--Pete
[right][snapback]87977[/snapback][/right]

Moody Blues, certainly. I don't recall the song title, though.

[LochnarITB]

"Of course you are, my bright little star . . ."

"I'm more than that. At least, I think I must be."

(Name that tune :) )

--Pete
[right][snapback]87977[/snapback][/right]

Heh, we've had this contest before. I think I was the one to post the lyrics that time, but I've grabbed them here.

Moody Blues

In The Beginning

[first man]
I think; I think I am; therefore I am; I think

[establishment]
Of course you are, my bright little star,
I've miles
And miles
Of files
Pretty files of your forefathers' fruit and now to suit our great computer
You're [sic] magnetic ink.

[first man]
I'm more than that; I know I am; at least, I think I must be.

[inner man]
There you go, man, keep as cool as you can,
Face piles
Of trials
With smiles.
It riles them to believe that you perceive the web they weave
And keep on thinking free.

[whyBish]

Vaguely-interesting tech story.

http://www.scienceblog.com/cms/node/8802
[right][snapback]87946[/snapback][/right]

Yeah, it sounds interesting in that it is generic, but the layperson needs to realise the difference between "grammatically correct" and "meaningful"