We propose a method to combine the interpretability and expressive power of firstorder logic with the effectiveness of neural network learning. In particular, we introduce a lifted framework in which first-order rules are used to describe the structure of a given problem setting. These rules are then used as a template for constructing a number of neural networks, one for each training and testing example. As the different networks corresponding to different examples share their weights, these weights can be efficiently learned using stochastic gradient descent. Our framework provides a flexible way for implementing and combining a wide variety of modelling constructs. In particular, the use of first-order logic allows for a declarative specification of latent relational structures, which can then be efficiently discovered in a given data set using neural network learning. Experiments on 78 relational learning benchmarks clearly demonstrate the effectiveness of the framework.

It is known due to the work of Van den Broeck, Meert and Darwiche that weighted first-order model counting (WFOMC) in the two-variable fragment of first-order logic can be solved in time polynomial in the number of domain elements. In this paper we extend this result to the two-variable fragment with counting quantifiers.

We describe an algorithm for constructing a set of tree-like conjunctive relational features by combining smaller conjunctive blocks. Unlike traditional level-wise approaches which preserve the monotonicity of frequency, our block-wise approach preserves monotonicity of feature reducibility and redundancy, which are important in propositionalization employed in the context of classification learning. With pruning based on these properties, our block-wise approach efficiently scales to features including tens of first-order atoms, far beyond the reach of state-of-the art propositionalization or inductive logic programming systems.

The field of Statistical Relational Learning (SRL) is concerned with learning probabilistic models from relational data. Learned SRL models are typically represented using some kind of weighted logical formulas, which make them considerably more interpretable than those obtained by e.g. neural networks. In practice, however, these models are often still difficult to interpret correctly, as they can contain many formulas that interact in non-trivial ways and weights do not always have an intuitive meaning. To address this, we propose a new SRL method which uses possibilistic logic to encode relational models. Learned models are then essentially stratified classical theories, which explicitly encode what can be derived with a given level of certainty. Compared to Markov Logic Networks (MLNs), our method is faster and produces considerably more interpretable models.

Abstract. Lifted relational neural networks (LRNNs) are a flexible neuralsymbolic framework based on the idea of lifted modelling. In this paper we show how LRNNs can be easily used to specify declaratively and solve learning problems in which latent categories of entities, properties and relations need to be jointly induced.