Alert button

QNLP in Practice: Running Compositional Models of Meaning on a Quantum Computer

Add code
Bookmark button
Alert button
Feb 25, 2021
Robin Lorenz, Anna Pearson, Konstantinos Meichanetzidis, Dimitri Kartsaklis, Bob Coecke

Figure 1 for QNLP in Practice: Running Compositional Models of Meaning on a Quantum Computer
Figure 2 for QNLP in Practice: Running Compositional Models of Meaning on a Quantum Computer
Figure 3 for QNLP in Practice: Running Compositional Models of Meaning on a Quantum Computer
Figure 4 for QNLP in Practice: Running Compositional Models of Meaning on a Quantum Computer

Share this with someone who'll enjoy it:

Twitter IconFacebook IconLinkedin IconWhatsapp IconMessenger Icon

Quantum Natural Language Processing (QNLP) deals with the design and implementation of NLP models intended to be run on quantum hardware. In this paper, we present results on the first NLP experiments conducted on Noisy Intermediate-Scale Quantum (NISQ) computers for datasets of size >= 100 sentences. Exploiting the formal similarity of the compositional model of meaning by Coecke et al. (2010) with quantum theory, we create representations for sentences that have a natural mapping to quantum circuits. We use these representations to implement and successfully train two NLP models that solve simple sentence classification tasks on quantum hardware. We describe in detail the main principles, the process and challenges of these experiments, in a way accessible to NLP researchers, thus paving the way for practical Quantum Natural Language Processing.

View paper onarxiv icon

Share this with someone who'll enjoy it:

Twitter IconFacebook IconLinkedin IconWhatsapp IconMessenger Icon