Qwiery is a symbolic cognitive architecture for building agents, text-based user interfaces and cognitive pipelines in general. It's a constuctive approach to cognitive computing and computational knowledge based on a modular system (plugins) with a focus on NLP-oriented interactions but allows a wide range of scenarios, including hybrid combinations of technologies (Python, R, Neo4j...) and approaches (neural networks, machine learning...).
Out of the box you will find in Qwiery a broad range of functionalities and features:
See the documentation for a complete feature list.
It should be emphasized that Qwiery does not exclude deep learning, supervised machine learning and reinforcement learning. The framework allows one to include deep learning models as interpreters and Qwiery can be used to train models through interactions. The plugin system and the agent pipelines are a way to organize and integrate multiple processes (possibly in parallel) with a slight emphasis on NLP, but you are free to inject things in whatever way you see fit.
npm i qwiery
In a JS file add
const Qwiery = require("qwiery"); Qwiery.repl();
and run it with
to see a basic REPL interface to Qwiery. There are many other ways you can use it, see the documentation for more.
Support and feedback