Skip to content
zdx3578 edited this page Oct 12, 2013 · 5 revisions

Provides a free utility for converting words into SDRs, and SDRs back into the closest matching word. Maintained by Francisco Webber, who's been active on our mailing list. Details about registering for an APP ID and KEY are on our mailing list. More endpoints are also available

Simple Python CEPT API client library.

From Chetan Surpur, details here.

From James Tauber, details here

[Offline word SDRS](offline retina mapping of words) for nupic-texts

If you don't want to get a CEPT API account, you can download word SDRs for the corpus of text output by nupic-texts in this convenient text file, which contains bitmap SDRs for all words for offline processing.

From Matt Taylor, extracts nouns from the NLTK corpus, converts them to plurals, and downloads SDRs from CEPT into a local cache.

[Human Like Machine Translator HLMT] (http://hlmt.net/)

Results were surprising and compared to other approaches such as MBMT and [PBMBMT] (http://ilk.uvt.nl/mbmt/pbmbmt/pbmbmt.pdf) has better results also compared to SMT

[details here] http://lists.numenta.org/pipermail/nupic_lists.numenta.org/2013-July/000655.html
http://hlmt.net

Clone this wiki locally