TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Rasa NLU: Open-source bot tool for natural language understanding

136 pointsby geospeckover 8 years ago

11 comments

espadrineover 8 years ago
For NLP, they use either MITIE[0] or spaCy[1].<p>That said, from my experience, you can get surprisingly far with simple systems; for instance, queread[2] relies on graph learning and statistics.<p>[0]: <a href="https:&#x2F;&#x2F;github.com&#x2F;mit-nlp&#x2F;MITIE" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;mit-nlp&#x2F;MITIE</a><p>[1]: <a href="https:&#x2F;&#x2F;spacy.io&#x2F;" rel="nofollow">https:&#x2F;&#x2F;spacy.io&#x2F;</a><p>[2]: <a href="https:&#x2F;&#x2F;github.com&#x2F;espadrine&#x2F;queread#workings" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;espadrine&#x2F;queread#workings</a>
IshKebabover 8 years ago
A while ago I looked for information on how Alexa, Wit.ai, Nuance Mix etc. do this intent classification and didn&#x27;t find anything.<p>These guys have posted a nice blog post about their approach:<p><a href="https:&#x2F;&#x2F;conversations.golastmile.com&#x2F;do-it-yourself-nlp-for-bot-developers-2e2da2817f3d#.l62loslgd" rel="nofollow">https:&#x2F;&#x2F;conversations.golastmile.com&#x2F;do-it-yourself-nlp-for-...</a><p>They suggest that they add the word vectors in the sentence. But it seems to me that that would make the result independent of the order of words (i.e. &quot;when does Tesco open?&quot; and &quot;Open Tesco when does&quot; are the same). I thought I had tested that and it didn&#x27;t work but actually I just tried saying &quot;Tesco open does when?&quot; to Alexa and it said &quot;Sorry, I don&#x27;t have the business hours for Tesco&quot;. Inconclusive I&#x27;d say but interesting anyway!
评论 #13206123 未加载
Rabidgremlinover 8 years ago
It&#x27;s the &quot;conversation&quot; part that is really tricky... I have been working on a bot for a large Corp for the last few months and we have been using Inkle&#x27;s Ink narration&#x2F;dialog engine for this. Works very well. They let me open source the framework: <a href="https:&#x2F;&#x2F;github.com&#x2F;rabidgremlin&#x2F;Mutters" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;rabidgremlin&#x2F;Mutters</a> it uses OpenNLP for intent identification and NER, Ink for conversation state and &quot;scripting&quot;
Maarten88over 8 years ago
This is interesting, I&#x27;ve been using LUIS for some time now and an open source alternative - especially one that is drop-in API compatible - is very welcome.<p>However I can&#x27;t find any information in the docs on how comparable the results are (i.e. does it have built-in date and time entity recognition like LUIS?). Most importantly: what languages does this support? All examples are in english-only. Is it even language aware, or do you train a model in any language? I&#x27;d be very interested if this were to support languages that LUIS does not have (like my language: Dutch)
评论 #13206074 未加载
评论 #13206134 未加载
ragebolover 8 years ago
Nice, I&#x27;ve been looking for an offline solution to do this sort of thing to run on a robot for RoboCup@Home.<p>Perhaps <a href="http:&#x2F;&#x2F;sag.art.uniroma2.it&#x2F;demo-software&#x2F;huric&#x2F;" rel="nofollow">http:&#x2F;&#x2F;sag.art.uniroma2.it&#x2F;demo-software&#x2F;huric&#x2F;</a> might also provide some training data. It&#x27;s annoying though I can&#x27;t just download that corpus but have to email some guy first.
niklasberover 8 years ago
Seems like it doesn&#x27;t say anywhere which language(s) it support? Guessing it&#x27;s English only.
评论 #13206005 未加载
mark_l_watsonover 8 years ago
Looks like an interesting project, based on skilearn and spaCy. The project provides some simple training files for the domain of asking about restaurants.<p>It would be useful to also have very large training data sets available.
nrp12over 8 years ago
Cool stuff - was looking for Open source NLU alternatives for luis.ai. Thanks to the emulators, this fits right in.<p>Does anyone know why rasa chose mitie&#x2F;spacy and not stanfordnlp?
评论 #13210016 未加载
qhocover 8 years ago
How well does this scale? Let&#x27;s say I have 500MB of JSON files from restaurant info and user reviews.
评论 #13209519 未加载
very_goordover 8 years ago
Great stuff, Is there a Docker support already ?
评论 #13206313 未加载
Coldeweyover 8 years ago
like the idea :-) good job!