search

IBM / LNN

A `Neural = Symbolic` framework for sound and complete weighted real-value logic
https://IBM.github.io/LNN/
Apache License 2.0
226 stars 438 forks source link

Can LNN generates logic rules/hypothesis from the input data (facts)? #52

Closed midoriririta closed 1 year ago

midoriririta commented 1 year ago

The given examples and tutorial shows LNN can do the logical reasoning from given data and root formulae. Can LNN generate logic hypothesis from the given data? Like, can it output the possible logic rules induced from background knowledges?

Thank you.

NaweedAghmad commented 1 year ago

Thank you for the question. Inductive logic programming is still an open problem that we and many others are looking at. Please have a look at the following paper to see how LNNs achieves interpretable, state of the art rules induced from data.

midoriririta commented 1 year ago

Hi thank you for your reply. I was wondering if I can find some code of example/demo of how to apply LNN on ilp task (with the induced rules as the output) in somewhere. It would be nice if you can provide a toy example of it. Thank you so much!