That does indeed sound very interesting Seyyed. Is there any chance we can
get access to this paper. If not can you predict on when it might be
available ?



On Fri, Jul 26, 2013 at 10:41 AM, Erik Blas <[email protected]> wrote:

> That sounds intriguing Seyyed!
>
>
> On Fri, Jul 26, 2013 at 12:08 AM, Seyyed Mohammad mohammadzadeh <
> [email protected]> wrote:
>
>> Hi,
>>
>> My MSC thesis was about using HTM/CLA to create a Machine Translator. The
>> work has been finished 2 weeks prior to the announcement of open-source CLA
>> so I've used my own implementation of CLA which was later improved, using
>> just the temporal pooler algorithm and incorporating it with an special
>> adaptive SDR generator. My new algorithm is called Adaptive Sequence Memory
>> (ASM) where the base structure and algorithms are based on CLA. Using ASM
>> I've created a Word Dictionary, a Multi-Word Dictionary and a Language
>> Model region. Combining them with an Incremental Learning Algorithm I could
>> create a new MT called Human Like Machine Translator 
>> (HLMT<http://hlmt.net/>).
>> Results were surprising and compared to other approaches such as 
>> MBMT<http://ilk.uvt.nl/mbmt/> and
>> PBMBMT <http://ilk.uvt.nl/mbmt/pbmbmt/pbmbmt.pdf> has better results
>> also compared to 
>> SMT<http://en.wikipedia.org/wiki/Statistical_machine_translation> it
>> uses much less resources increasing learning and translation speed while
>> keeping a competitive translation output. I hope my paper is accepted to be
>> presented in EMNLP 2013 <http://hum.csse.unimelb.edu.au/emnlp2013/> so I
>> can explain more about it.
>>
>>
>> 2013/7/25 James Tauber <[email protected]>
>>
>>>  So it turns out there *is* some interest. I've CC'd my old google
>>> group to try to rekindle things.
>>>
>>> James
>>>
>>>
>>> On Wed, Jul 24, 2013 at 12:50 PM, James Tauber <[email protected]>wrote:
>>>
>>>> Back when Jeff's book first came out and Numenta was founded, I started
>>>> a mailing list htm-ling for talking about HTM as applied to linguistics.
>>>>
>>>> The list never really had much conversation on it but I wondered if,
>>>> with NuPIC and the CLA, there might be renewed interest or existing work on
>>>> linguistic and NLP applications.
>>>>
>>>> Anyone interested or know of any work?
>>>>
>>>> Anyone at OSCON interested in talking about this more?
>>>>
>>>> James
>>>> --
>>>> James Tauber
>>>> http://jtauber.com/
>>>> @jtauber on Twitter
>>>>
>>>
>>>
>>>
>>> --
>>> James Tauber
>>> http://jtauber.com/
>>> @jtauber on Twitter
>>>
>>> _______________________________________________
>>> nupic mailing list
>>> [email protected]
>>> http://lists.numenta.org/mailman/listinfo/nupic_lists.numenta.org
>>>
>>>
>>
>>
>> --
>> Best Regards.
>>  S.M. Mohammadzadeh
>> *Web: www.mohammadzadeh.info*
>> *Blog: blog.mohammadzadeh.info*
>>
>> * softnhard ==> software and hardware expert
>>
>> _______________________________________________
>> nupic mailing list
>> [email protected]
>> http://lists.numenta.org/mailman/listinfo/nupic_lists.numenta.org
>>
>>
>
> _______________________________________________
> nupic mailing list
> [email protected]
> http://lists.numenta.org/mailman/listinfo/nupic_lists.numenta.org
>
>
_______________________________________________
nupic mailing list
[email protected]
http://lists.numenta.org/mailman/listinfo/nupic_lists.numenta.org

Reply via email to