Open-source intent recognition in NLP & NLU

Natural-language processing (short „NLP“) is an uprising area in the face of artificial intelligence. When building semi-intelligent systems, NLP tries to help developers to understand their users / customers / datasources (this is when your start talking about „Natural language understanding“ or NLU – a subtopic of natural language processing) . Especially intent or activity recognition is quite relevant to „convert“ spoken words into processable data. Even though this technical area is heating up, the eco-system of Open-source technology around it seems quite chaotic. This listing should provide a quick overview about some of the current Open-source frameworks and technologies available for intent / activity recognition:

Rasa NLU MycroftAI Adapt Open Intent spaCy
Rasa NLU (Natural Language Understanding) is a tool for understanding what is being said in short pieces of text. The Adapt Intent Parser is a flexible and extensible intent definition and determination framework. It is intended to parse natural language text into a structured intent that can then be invoked programatically. Open-intent will help you translate your user intents into actual business actions on messenger and your own website. spaCy is a free open-source library featuring state-of-the-art speed and accuracy and a powerful Python API
– Can be integrated via HTTP API

– Seems pretty major and stable (in comparison)

– Can use spaCy as a backend

– Seems pretty customizable, but a bit rough-edged

Node.js wrapper „AdaptJS“ available

– They seem to be building something new called „Padatious“

 – Comes with an own Open-Intent Markup Language – NLU is not provided by default

– They promote their proprietary tool „Prodigy“ for NLU

Those Open-source technologies provide a suitable alternative to the proprietary competitors wit,,, LUIS from Microsoft or Dialogflow from Google.