discopy 0.1.6

Distributional Compositional Python

Distributional Compositional Python

discopy is a tool for classical and quantum natural language processing (QNLP) using string diagrams.

from discopy import Ty, Word, Cup, Id

s, n = Ty('s'), Ty('n')
Alice, Bob = Word('Alice', n), Word('Bob', n)
loves = Word('loves', n.r @ s @ n.l)

sentence = Alice @ loves @ Bob >> Cup(n, n.r) @ Id(s) @ Cup(n.l, n)


# 2) Define a model.

from discopy import Model

ob = {s: 1, n: 2}
ar = {Alice: [1, 0], loves: [0, 1, 1, 0], Bob: [0, 1]}
F = Model(ob, ar)


# 3) Compute the meaning!

assert F(sentence)

Requirements

• numpy
• networkx
• matplotlib
• pytket

Getting Started

pip install discopy

Documentation

The documentation is hosted at readthedocs.io, you can also checkout the notebooks for a demo!

References

• Pregroup grammars and categorical compositional distributional semantics on the nLab
• From Word to Sentence: A Computational Algebraic Approach to Grammar - Lambek (2008)
• A Compositional Distributional Model of Meaning - Clark, Coecke, Sadrzadeh (2008)
• Experimental Support for a Categorical Compositional Distributional Model of Meaning - Grefenstette and Sadrzadeh (2010)
• Functorial Question Answering - De Felice, Meichanetzidis, Toumi (2019)
• The Mathematics of Text Structure - Coecke (2019)