LYRICS: A General Interface Layer to Integrate Logic Inference and Deep Learning

Marra, Giuseppe; Giannini, Francesco; Diligenti, Michelangelo; Gori, Marco

doi:10.1007/978-3-030-46147-8_17

Cited by 26 publications

(30 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Silvestri et al [57] shows how adding domain knowledge in the form of symbolic constraints greatly improves the samplingfrequency of a neural network trained to solve a combinatorial problem. The LYRICS system [42] proposes a generic interface layer that allows to define arbitrary first order logic background knowledge, allowing a learning system to learn its weights under the constraints imposed by the prior knowledge.…”

Section: Informed Learning With Prior Knowledgementioning

confidence: 99%

Modular design patterns for hybrid learning and reasoning systems

et al. 2021

View full text Add to dashboard Cite

The unification of statistical (data-driven) and symbolic (knowledge-driven) methods is widely recognized as one of the key challenges of modern AI. Recent years have seen a large number of publications on such hybrid neuro-symbolic AI systems. That rapidly growing literature is highly diverse, mostly empirical, and is lacking a unifying view of the large variety of these hybrid systems. In this paper, we analyze a large body of recent literature and we propose a set of modular design patterns for such hybrid, neuro-symbolic systems. We are able to describe the architecture of a very large number of hybrid systems by composing only a small set of elementary patterns as building blocks. The main contributions of this paper are: 1) a taxonomically organised vocabulary to describe both processes and data structures used in hybrid systems; 2) a set of 15+ design patterns for hybrid AI systems organized in a set of elementary patterns and a set of compositional patterns; 3) an application of these design patterns in two realistic use-cases for hybrid AI systems. Our patterns reveal similarities between systems that were not recognized until now. Finally, our design patterns extend and refine Kautz’s earlier attempt at categorizing neuro-symbolic architectures.

show abstract

Section: Informed Learning With Prior Knowledgementioning

confidence: 99%

Modular design patterns for hybrid learning and reasoning systems

et al. 2021

View full text Add to dashboard Cite

show abstract

“…[62] shows how adding domain knowledge in the form of symbolic constraints greatly improves the sampling-frequency of a neural network trained to solve a combinatorial problem. The LYRICS system [47] proposes a generic interface layer that allows to define arbitrary first order logic background knowledge, allowing a learning system to learn its weights under the constraints imposed by the prior knowledge.…”

Section: Informed Learning With Prior Knowledgementioning

confidence: 99%

Modular Design Patterns for Hybrid Learning and Reasoning Systems: a taxonomy, patterns and use cases

van Bekkum,

de Boer,

van Harmelen

et al. 2021

Preprint

View full text Add to dashboard Cite

The unification of statistical (data-driven) and symbolic (knowledge-driven) methods is widely recognised as one of the key challenges of modern AI. Recent years have seen large number of publications on such hybrid neuro-symbolic AI systems. That rapidly growing literature is highly diverse and mostly empirical, and is lacking a unifying view of the large variety of these hybrid systems. In this paper we analyse a large body of recent literature and we propose a set of modular design patterns for such hybrid, neuro-symbolic systems. We are able to describe the architecture of a very large number of hybrid systems by composing only a small set of elementary patterns as building blocks. The main contributions of this paper are: 1) a taxonomically organised vocabulary to describe both processes and data structures used in hybrid systems; 2) a set of 15+ design patterns for hybrid AI systems, organised in a set of elementary patterns and a set of compositional patterns; 3) an application of these design patterns in two realistic use-cases for hybrid AI systems. Our patterns reveal similarities between systems that were not recognised until now. Finally, our design patterns extend and refine Kautz' earlier attempt at categorising neuro-symbolic architectures.

show abstract

“…For instance, one can cast constants to vectors, functions terms to vector functions of the corresponding dimensionality, and similarly predicates to tensors of the corresponding aritydimension (Rocktäschel et al 2015;Diligenti et al 2017). Again adopting a fuzzy logic interpretation of the logical connectives, the learning problem can then be cast as a constrained numerical optimization problem, including works such as Logic Tensor Networks (Serafini and d'Avila Garcez 2016) or LYRICS (Marra et al 2019). While the distributed representation of the logical constructs is the subject of learning, in contrast with the discussed Datalog program structure learning approaches, the weight (strength) of each rule needs to be specified apriori --a limitation which was recently addressed in Marra et al (2020).…”

Section: Related Workmentioning

confidence: 99%

Beyond graph neural networks with lifted relational neural networks

2021

View full text Add to dashboard Cite

We introduce a declarative differentiable programming framework, based on the language of Lifted Relational Neural Networks, where small parameterized logic programs are used to encode deep relational learning scenarios through the underlying symmetries. When presented with relational data, such as various forms of graphs, the logic program interpreter dynamically unfolds differentiable computation graphs to be used for the program parameter optimization by standard means. Following from the declarative, relational logic-based encoding, this results into a unified representation of a wide range of neural models in the form of compact and elegant learning programs, in contrast to the existing procedural approaches operating directly on the computational graph level. We illustrate how this idea can be used for a concise encoding of existing advanced neural architectures, with the main focus on Graph Neural Networks (GNNs). Importantly, using the framework, we also show how the contemporary GNN models can be easily extended towards higher expressiveness in various ways. In the experiments, we demonstrate correctness and computation efficiency through comparison against specialized GNN frameworks, while shedding some light on the learning performance of the existing GNN models.

show abstract

LYRICS: A General Interface Layer to Integrate Logic Inference and Deep Learning

Cited by 26 publications

References 13 publications

Modular design patterns for hybrid learning and reasoning systems

Modular design patterns for hybrid learning and reasoning systems

Modular Design Patterns for Hybrid Learning and Reasoning Systems: a taxonomy, patterns and use cases

Beyond graph neural networks with lifted relational neural networks

Contact Info

Product

Resources

About