6

I recently discovered Cambridge Quantum people have developed lambeq, a quantum natural language processing high-level library. Before diving into it, I'd like to understand more in detail what quantum computers can do better when it comes to NLP. Essentially, the intuition of why one might want to use a quantum computer for NLP instead of a classical one.

I read the lambeq paper but they mainly focus on the design of the library, introducing QNLP as just a possible application of quantum computing.

mpro
  • 517
  • 2
  • 11
  • 1
    NLP and QNLP essentially deal with the same objects. But classical data requires exponentially more storage than quantum data, if encoded in amplitude. The DiscoCat model uses a graphical logic to represent semantics, which is very meaningful. But a series of papers on DiscoCat would point out that the model also requires a lot of storage. The problems that quantum computers can solve are very limited. Often papers promise quantum supremacy in advance, but the actual effect is unknown. – Ren-Xin Zhao Jul 18 '22 at 10:32
  • @R-XZhao Thanks for you answer. Can you provide more details (or redirect me to some resource or reference) to understand more the storage advantages you mentioned ? Thanks – mpro Jul 19 '22 at 07:10
  • Bob Coecke, Giovanni De Felice et al., “Foundations for near-term quantum natural language processing,” arXiv e-prints, pp. arXiv:2012.03755, 2020. – Ren-Xin Zhao Jul 19 '22 at 09:22
  • Thanks a lot, that's exactly the kind of resource I was looking for. If you put it as an answer to my question (not as a comment), I will accept it. – mpro Jul 19 '22 at 09:56

1 Answers1

1

There is no doubt that the DiscoCat model is novel. However, the generalization of the advantages of quantum computers has not yet been realized. So there's a good chance you'll get a less-than-classical effect. Please continue to refer to

  1. Bob Coecke, Giovanni De Felice et al., “Foundations for near-term quantum natural language processing,” arXiv e-prints, pp. arXiv:2012.03755, 2020.
  2. Robin Lorenz, Anna Pearson et al., “Qnlp in practice: Running compositional models of meaning on a quantum computer,” arXiv e-prints, pp. arXiv:2102.12846, 2021.

You can follow the Google Scholar feed of Cambridge University professor Bob Coecke, but there is very little material on QNLP at the moment.

Ren-Xin Zhao
  • 536
  • 3
  • 12