New algorithm for learning languages
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.
The development -- which has a patent pending -- has implications for speech recognition and for other applications in natural language engineering, as well as for genomics and proteomics. It also offers new insights into language acquisition and psycholinguistics.
"The algorithm -- the computational method -- for language learning and processing that we have developed can take a body of text, abstract from it a collection of recurring patterns or rules and then generate new material," explained Shimon Edelman, a computer scientist who is a professor of psychology at Cornell and co-author of a new paper, "Unsupervised Learning of Natural Languages," published in the Proceedings of the National Academy of Sciences (PNAS, Vol. 102, No. 33).
"This is the first time an unsupervised algorithm is shown capable of learning complex syntax, generating grammatical new sentences and proving useful in other fields that call for structure discovery from raw data, such as bioinformatics," he said.
Unlike previous attempts at developing computer algorithms for language learning, the new method, called Automatic Distillation of Structure (ADIOS), successfully identifies complex patterns in raw texts. The algorithm discovers the patterns by repeatedly aligning sentences and looking for overlapping parts.
For example, the sentences I would like to book a first-class flight to Chicago, I want to book a first-class flight to Boston and Book a first-class flight for me, please may give rise to the pattern book a first-class flight -- if this candidate pattern passes the novel statistical significance test that is the core of the algorithm.
If the system also encounters the sentences I need to book a direct flight from New York to Tel Aviv andI would like to book an economy flight , it may infer that the phrases first-class, direct and economy are equivalent in the context of the new pattern. "Because such equivalence sets can contain other patterns -- in turn containing further patterns, and so on -- the resulting body of knowledge grows recursively, as a sort of forest of branching trees of possibilities," said Edelman.
He added, "ADIOS relies on a statistical method for pattern extraction and on structured generalization -- two processes that have been implicated in language acquisition. Our experiments show that it can acquire intricate structures from raw data, including transcripts of parents' speech directed at 2- or 3-year-olds. This may eventually help researchers understand how children, who learn language in a similar item-by-item fashion and with very little supervision, eventually master the full complexities of their native tongue."
In addition to child-directed language, the algorithm has been tested on the full text of the Bible in several languages, on artificial context-free languages with thousands of rules and on musical notation. It also has been applied to biological data, such as nucleotide base pairs and amino acid sequences. In analyzing proteins, for example, the algorithm was able to extract from amino acid sequences patterns that were highly correlated with the functional properties of the proteins.
The new method was developed jointly with David Horn and Eytan Ruppin, professors of physics and computer science, respectively, at Tel Aviv University, and with Zach Solan, a doctoral student there and the lead author on the paper. Their collaboration with Edelman was supported in part by the U.S.-Israel Binational Science Foundation.
From Cornell University
- 26675 reads
Printer-friendly version | 
Send to friend
FwygtzlXKM
ivaedocvunf@gmail.com
Great
Really its wonderful to create such algo to understand any language .I appreciate your work.
Hope, yes
Hope, yes
Madrid student
This article was of great information. I came across various rules and regulation which could later generate new and meaningful sentences in future.
business english course
First of all thank you for posting an article on this topic and secondly good luck on the quick start of new algorithm for learning languages. I wonder if you could discover more on business english course or check out the site http://www.british-study.com/adults/english-language-courses/business-english.php
I would like to book a
I would like to book a first-class flight to Chicago, I want to book a first-class flight to Boston and Book a first-class flight for me, please may give rise to the pattern book a first-class flight -- if this candidate pattern passes the novel statistical significance test that is the core of the algorithm.:)
Full Songs
Lida going Lida
Lida going Lida
Thanks
Thanks for sharing these info with us! I was reading something similar on another website that i was researching. I will be sure to look around more. thanks...
Thanks
First, thanks for new algorithm, then,how this algorithm is different? i will try..
~~~~~~~~
james
msn cams
Sheet music
I like the idea of creating meaningful, yet random sheet music.
k?zl?k bozma
It's an old idea (if I don't miss something). If you are interested in this topic use Google and search for "Markov chains". http://www.xhamsterturk.com Spammers generate "like-real" tests using this algorithm. And yes, it also can be applied to test texts.
backup
I make backups http://rapid4me.com/?q=backup of my kids movies so that WHEN they trash them (I gave up on thinking “if”) I don’t have to spill another $30 because it was a Disney flick. I already bought Mulan and Toy Story twice.. I learned my lesson.
fun game
some Gaming is so fun,and i like thecosplay for mmo too
Thanks for sharing these info
Thanks for sharing these info with us! I was reading something similar on another website that i was researching. I will be sure to look around more. thanks...
thank you
thanks
Yusuf Güney
r10.net genclik^^
The method also works for such data as sheet music or protein sequences.
Excellently written article,
Excellently written article, if only all bloggers offered the same content as you, the internet would be a much better place. Please keep it up!
Cheers.
I study
I study Applied Linguistics in my university, Computational and other lingustics. My purpose is to.
Yeah!
Yeah, thats fantastic, let it be.
Yeah!
Yeah, thats fantastic, let it be.
Great
Yes, it's great feature for spam prevention development.
Sure
It's really fantastic.
Re: spam?
I heard about that algorithm too.
What is the difference?
Hello, how this algorithm is different?
Really can improve my study of any languaje know this?
Regards,
Vivi
http://www.deperfumes.com
Why
What do it ? video ?
1tk
Why spam a the real security http://www.1tk.org
)
Is it being actually tested to work?
http://www.casininio.com
Wow that is a fantastic
Wow that is a fantastic development, it should be able to be modified to combat spam surely?
It's an old idea (if I don't
It's an old idea (if I don't miss something). If you are interested in this topic use Google and search for "Markov chains". Spammers generate "like-real" tests using this algorithm. And yes, it also can be applied to test texts.
mobile forum
Goof example
For example, the sentences I would like to book a first-class flight to Chicago, I want to book a first-class flight to Boston and Book a first-class flight for me, please may give rise to the pattern book a first-class flight -- if this candidate pattern passes the novel statistical significance test that is the core of the algorithm.:)
____
Aron
Thanks
thanks 4 post
___________________
aytug akdogan ödüllü 1. seo yar??mas? ve yurtta bar?? dünyada bar??
Fenerbahçe
Lida
Bro
I need turkish language please.
aytug akdogan ödüllü 1. seo yar??mas? ve yurtta bar?? dünyada bar??
there is always the problem
there is always the problem of 'how much is too much'? as you said it is applied on extensive and recurring raw data. where do you get this data from? i don't really care, but what about the validity of the data? in order to get a correct algorithm you need correct data.
Strict pattern-based methods
Strict pattern-based methods of grammar induction are often frustrated by the apparently inexhaustible variety of novel word combinations in large corpora. Statistical methods offer a possible solution by allowing frequent well-formed expressions to overwhelm the infrequent ungrammatical ones. They also have the desirable property of being able to construct robust grammars from positive instances alone. Unfortunately, the zero-frequency problem entails assigning a small probability to all possible word patterns, thus ungrammatical n-grams become as probable as unseen grammatical ones. Further, such grammars are unable to take advantage of inherent lexical properties that should allow infrequent words to inherit the syntactic properties of the class to which they belong.
Yes it's pity but it is
Yes it's pity but it is so:(
Search engines become smarter only because spa*mers getting smarter and invent such linguistic tools.
It's an old idea (if I don't
It's an old idea (if I don't miss something). If you are interested in this topic use Google and search for "Markov chains". Spammers generate "like-real" tests using this algorithm. And yes, it also can be applied to test texts.
--
Treat
We create program
create program , that can easily generate texts with content, that answers exactly to human's questions. I think it'll be possible, when smth about background knowledge will be created.
I study
I study Applied Linguistics in my university, Computational and other lingustics. My purpose is to
Well this is interesting
Well this is interesting I imagine the algorithm is in some version of C+ or something. There was once three or four competing lists of rules for English syntax, about 33 or so, so this is a good thing for computers to do, so to speak, sort of a real-time syntactical concordance analyses.
Syntax
I once studied syntax and transformational grammar at Stony Brook University, along with other linguistic classes for Anthropology and find this an interesting development. When computer languages got started there was one SNOBOL which processed language, instead of numbers, which I thought might some day be developed, why even Bill Gates once promised a SNOBOL for Windows. (Where is it Mr. Gates?) Well this is interesting I imagine the algorithm is in some version of C+ or something. There was once three or four competing lists of rules for English syntax, about 33 or so, so this is a good thing for computers to do, so to speak, sort of a real-time syntactical concordance analyses. Bravo!
George J Myers, Jr. (my first post here, I'm awed)
I need a French/English
I need a French/English version now please...before it's too late.
or even better, a spam
or even better, a spam generator!
-bugmenot.com-
spam?
Since spam often consists of randomly generated sentences could this algorithm be used as a spam filter?
Post new comment