Help to with NLP in NuPIC

[baymaxx]

I have a dataset with many sentences. I want to send them to the temporal memory algorithm and so when the next time, a part of a sentence is given as input, I want the algorithm to predict the next word.

I have seen the demo of text prediction in HTM school videos. But, I would like to know how I can do it.

[marty1885]

I have done the same thing before! See

Esperanto NLP using HTM and my findings

Hi all. I have been messing with NLP using Temporal Memory in the weekends. I coded up a simple program with and made it to learn the structure of Esperanto sentences. (I’m testing with Esperanto because the language itself is regular and has obvious features.) And my results are… not surprising. HTM isn’t the most powerful ML model yet and it performs close-to or slightly better then a RNN. Yet it is still powerful in the sense that there is no need to backprop all the way trough the entire s…

[baymaxx]

Hey, thanks for the reply. I went through your blog and it was useful. But could I get some info on developing good predicting NLP models (predicting words) given part of a sentence? Maybe I could use Cortical.IO? I just started to learn about NuPIC and NLP.

[rhyolight]

This is quite old, but potentially relevant. I have always thought that encoding POS could really help.

This is a hackathon kickoff video includes some examples of using NuPIC to predict POS.