Our work is focused on compositionality, and we have developed ZX-calculus, a low-level diagrammatic language for reasoning about quantum computations. As we employ the ZX-calculus to Quantum Natural Language Processing (QNLP) architectures, we continue to advance the theory of graphical languages.
We have implemented proof-of-concept experiments, solving NLP tasks such as sentence classification using techniques from the rapidly developing field of Quantum Machine Learning to automatically learn parameterized “language circuits”.
Computational linguistics is a field that strives to understand language from a computational perspective. A computer that could achieve human-like facility with language would theoretically be able to engage in meaningful dialogue; acquire new languages and translate among them; and increase its knowledge by “reading” text. We believe quantum computers are particularly suited to such a field of study.
A computational understanding of language might also provide insight into how the human brain itself functions. Through computational linguistics, we seek to create a linguistically competent computer and deepen our understanding of the way we think.
quantum theory and natural languages
LAMBEQ was created to explore the structural correspondences between the matching compositional structures underpinning quantum theory, structures such as process theories and tensor networks, with formal and natural languages to design novel models for language and other cognitive phenomena.
An open-source software library, LAMBEQ, enables the design and implementation of end-to-end QNLP pipelines that seamlessly integrate with Quantinuum’s platform-agnostic compiler, TKET.
Prof. Bob Coecke, Chief Scientist, leads Quantinuum’s Quantum Compositional Intelligence and Computational Linguistics teams.
Join the Quantum Compositional Intelligence Research Ecosystem and explore the structural relationships between quantum theory and natural languages.