Discrete Word Embedding for Logical Natural Language Understanding

Zilu Tang, Masataro Asai

Keywords: NLP, word embedding, discrete VAE, classical planning, neural symbolic

Abstract:

We propose an unsupervised neural model for learning a discrete embedding of words.
Unlike existing discrete embeddings, our binary embedding supports vector arithmetic operations similar to continuous embeddings.
Our embedding represents each word as a set of propositional statements describing a transition rule in classical/STRIPS planning formalism.
This makes the embedding directly compatible with symbolic, state of the art classical planning solvers.

One-sentence Summary: We propose an unsupervised neural model for learning a discrete embedding of words compatible with classical planning solvers.

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics

Reviewed Version (pdf): https://openreview.net/references/pdf?id=l4v0zevmLl