Skip to main navigation Skip to search Skip to main content

Inference in probabilistic logic programs using lifted explanations

  • Stony Brook University

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

In this paper, we consider the problem of lifted inference in the context of Prism-like probabilistic logic programming languages. Traditional inference in such languages involves the construction of an explanation graph for the query that treats each instance of a random variable separately. For many programs and queries, we observe that explanations can be summarized into substantially more compact structures introduced in this paper, called "lifted explanation graphs". In contrast to existing lifted inference techniques, our method for constructing lifted explanations naturally generalizes existing methods for constructing explanation graphs. To compute probability of query answers, we solve recurrences generated from the lifted graphs. We show examples where the use of our technique reduces the asymptotic complexity of inference.

Original languageEnglish
Title of host publicationTechnical Communications of the 32nd International Conference on Logic Programming, ICLP 2016
EditorsAndy King, Neda Saeedloei, Marina De Vos, Manuel Carro Linares
PublisherSchloss Dagstuhl- Leibniz-Zentrum fur Informatik GmbH, Dagstuhl Publishing
ISBN (Electronic)9783959770071
DOIs
StatePublished - Nov 1 2016
Event32nd International Conference on Logic Programming, ICLP 2016 - New York City, United States
Duration: Oct 16 2016Oct 21 2016

Publication series

NameOpenAccess Series in Informatics
Volume52
ISSN (Print)2190-6807

Conference

Conference32nd International Conference on Logic Programming, ICLP 2016
Country/TerritoryUnited States
CityNew York City
Period10/16/1610/21/16

Keywords

  • Constraints
  • Lifted inference
  • Probabilistic inference
  • Probabilistic logic programs
  • Symbolic evaluation

Fingerprint

Dive into the research topics of 'Inference in probabilistic logic programs using lifted explanations'. Together they form a unique fingerprint.

Cite this