Implementation and performance of probabilistic inference pipelines

Dimitar Shterionov*, Gerda Janssens

*Corresponding author for this work

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

5 Citations (Scopus)


In order to handle real-world problems, state-of-the-art probabilistic logic and learning frameworks, such as ProbLog, reduce the expensive inference to an efficient Weighted Model Counting. To do so ProbLog employs a sequence of transformation steps, called an inference pipeline. Each step in the probabilistic inference pipeline is called a pipeline component. The choice of the mechanism to implement a component can be crucial to the performance of the system. In this paper we describe in detail different ProbLog pipelines. Then we perform a empirical analysis to determine which components have a crucial impact on the efficiency. Our results show that the Boolean formula conversion is the crucial component in an inference pipeline. Our main contributions are the thorough analysis of ProbLog inference pipelines and the introduction of new pipelines, one of which performs very well on our benchmarks.

Original languageEnglish
Title of host publicationPractical Aspects of Declarative Languages - 17th International Symposium, PADL 2015, Proceedings
EditorsTran Cao Son, Enrico Pontelli
PublisherSpringer Verlag
Number of pages15
ISBN (Print)9783319196855
Publication statusPublished - 2015
Externally publishedYes
Event17th International Symposium on Practical Aspects of Declarative Languages, PADL 2015 - Portland, United States
Duration: 18 Jun 201519 Jun 2015

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349


Conference17th International Symposium on Practical Aspects of Declarative Languages, PADL 2015
Country/TerritoryUnited States


Dive into the research topics of 'Implementation and performance of probabilistic inference pipelines'. Together they form a unique fingerprint.

Cite this