Scalable knowledge base completion with superposition memories

10/24/2021
by   Matthias Lalisse, et al.
8

We present Harmonic Memory Networks (HMem), a neural architecture for knowledge base completion that models entities as weighted sums of pairwise bindings between an entity's neighbors and corresponding relations. Since entities are modeled as aggregated neighborhoods, representations of unseen entities can be generated on the fly. We demonstrate this with two new datasets: WNGen and FBGen. Experiments show that the model is SOTA on benchmarks, and flexible enough to evolve without retraining as the knowledge graph grows.

READ FULL TEXT

Please sign up or login with your details

Forgot password? Click here to reset