Neural Markov Logic Networks
This work addresses the challenge of automating rule learning in relational models for researchers and practitioners in AI, offering a novel hybrid approach that combines neural and symbolic methods.
The paper tackles the problem of statistical relational learning by introducing neural Markov logic networks (NMLNs), which learn implicit first-order logic rules as neural network potentials, eliminating the need for explicit rule specification. The result is a system that performs well in knowledge-base completion, triple classification, and molecular graph data generation, even without constant embeddings in non-transductive settings.
We introduce neural Markov logic networks (NMLNs), a statistical relational learning system that borrows ideas from Markov logic. Like Markov logic networks (MLNs), NMLNs are an exponential-family model for modelling distributions over possible worlds, but unlike MLNs, they do not rely on explicitly specified first-order logic rules. Instead, NMLNs learn an implicit representation of such rules as a neural network that acts as a potential function on fragments of the relational structure. Similarly to many neural symbolic methods, NMLNs can exploit embeddings of constants but, unlike them, NMLNs work well also in their absence. This is extremely important for predicting in settings other than the transductive one. We showcase the potential of NMLNs on knowledge-base completion, triple classification and on generation of molecular (graph) data.