DNF Sampling for ProbLog Inference

TL;DR

This paper introduces a novel DNF sampling approximation method for ProbLog inference, leveraging the DNF structure to improve sampling efficiency, and compares it experimentally with existing methods.

Contribution

It presents the first application of DNF sampling to probabilistic logic inference, offering a new approach to approximate ProbLog inference.

Findings

The DNF sampling method effectively approximates ProbLog inference.

Experimental results show competitive performance against existing sampling methods.

The approach reduces computational complexity in probabilistic logic inference.

Abstract

Inference in probabilistic logic languages such as ProbLog, an extension of Prolog with probabilistic facts, is often based on a reduction to a propositional formula in DNF. Calculating the probability of such a formula involves the disjoint-sum-problem, which is computationally hard. In this work we introduce a new approximation method for ProbLog inference which exploits the DNF to focus sampling. While this DNF sampling technique has been applied to a variety of tasks before, to the best of our knowledge it has not been used for inference in probabilistic logic systems. The paper also presents an experimental comparison with another sampling based inference method previously introduced for ProbLog.