Semirings for probabilistic and neuro-symbolic logic programming

The field of probabilistic logic programming (PLP) focuses on integrating probabilistic models into programming languages based on logic. Over the past 30 years, numerous languages and frameworks have been developed for modeling, inference and learning in probabilistic logic programs. While original...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:International journal of approximate reasoning Jg. 171; S. 109130
Hauptverfasser: Derkinderen, Vincent, Manhaeve, Robin, Zuidberg Dos Martires, Pedro, De Raedt, Luc
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Elsevier Inc 01.08.2024
Schlagworte:
Model counting
Neuro-symbolic AI
Probabilistic logic programming
Semiring programming
Semiring programming
Model counting
Neuro-symbolic AI
Probabilistic logic programming
ISSN:0888-613X, 1873-4731, 1873-4731
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:The field of probabilistic logic programming (PLP) focuses on integrating probabilistic models into programming languages based on logic. Over the past 30 years, numerous languages and frameworks have been developed for modeling, inference and learning in probabilistic logic programs. While originally PLP focused on discrete probability, more recent approaches have incorporated continuous distributions as well as neural networks, effectively yielding neuro-symbolic methods. We provide an overview and synthesis of this domain, thereby contributing a unified algebraic perspective on the different flavors of PLP, showing that many if not most of the extensions of PLP can be cast within a common algebraic logic programming framework, in which facts are labeled with elements of a semiring and disjunction and conjunction are replaced by addition and multiplication. This does not only hold for the PLP variations itself but also for the underlying execution mechanism that is based on (algebraic) model counting. In order to showcase and explain this unified perspective, we focus on the ProbLog language and its extensions.
ISSN:0888-613X
1873-4731
1873-4731
DOI:10.1016/j.ijar.2024.109130