DNF sampling for ProbLog inference

Dimitar S. Shterionov; Angelika Kimmig; Theofrastos Mantadelis; Gerda Janssens

DNF sampling for ProbLog inference

Dimitar S. Shterionov
, Angelika Kimmig
, Theofrastos Mantadelis
, Gerda Janssens

Research output: Contribution to conference › Paper › Other research output

Abstract

Inference in probabilistic logic languages such as ProbLog, an extension of Prolog with probabilistic facts, is often based on a reduc- tion to a propositional formula in DNF. Calculating the probability of such a formula involves the disjoint-sum-problem, which is computation- ally 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 probabilis- tic logic systems. The paper also presents an experimental comparison with another sampling based inference method previously introduced for ProbLog.

Original language	English
Number of pages	1
Publication status	Published - 2010
Externally published	Yes
Event	26th International Conference on Logic Programming, ICLP 2010 - Joint Workshop on Implementation of Constraint Logic Programming Systems and Logic-based Methods in Programming Environments - CICLOPS-WLPE 2010 - Edinburgh, Scotland, United Kingdom Duration: 15 Jul 2010 → 15 Jul 2010

Conference

Conference	26th International Conference on Logic Programming, ICLP 2010 - Joint Workshop on Implementation of Constraint Logic Programming Systems and Logic-based Methods in Programming Environments - CICLOPS-WLPE 2010
Country/Territory	United Kingdom
City	Edinburgh, Scotland
Period	15/07/10 → 15/07/10

Cite this

Shterionov, D. S., Kimmig, A., Mantadelis, T., & Janssens, G. (2010). DNF sampling for ProbLog inference. Paper presented at 26th International Conference on Logic Programming, ICLP 2010 - Joint Workshop on Implementation of Constraint Logic Programming Systems and Logic-based Methods in Programming Environments - CICLOPS-WLPE 2010, Edinburgh, Scotland, United Kingdom.

Shterionov, Dimitar S. ; Kimmig, Angelika ; Mantadelis, Theofrastos et al. / DNF sampling for ProbLog inference. Paper presented at 26th International Conference on Logic Programming, ICLP 2010 - Joint Workshop on Implementation of Constraint Logic Programming Systems and Logic-based Methods in Programming Environments - CICLOPS-WLPE 2010, Edinburgh, Scotland, United Kingdom.1 p.

@conference{ba89147ec94b49bf8472d9a10ce3edea,

title = "DNF sampling for ProbLog inference",

abstract = "Inference in probabilistic logic languages such as ProbLog, an extension of Prolog with probabilistic facts, is often based on a reduc- tion to a propositional formula in DNF. Calculating the probability of such a formula involves the disjoint-sum-problem, which is computation- ally 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 probabilis- tic logic systems. The paper also presents an experimental comparison with another sampling based inference method previously introduced for ProbLog.",

author = "Shterionov, \{Dimitar S.\} and Angelika Kimmig and Theofrastos Mantadelis and Gerda Janssens",

year = "2010",

language = "English",

note = "26th International Conference on Logic Programming, ICLP 2010 - Joint Workshop on Implementation of Constraint Logic Programming Systems and Logic-based Methods in Programming Environments - CICLOPS-WLPE 2010 ; Conference date: 15-07-2010 Through 15-07-2010",

}

Shterionov, DS, Kimmig, A, Mantadelis, T & Janssens, G 2010, 'DNF sampling for ProbLog inference', Paper presented at 26th International Conference on Logic Programming, ICLP 2010 - Joint Workshop on Implementation of Constraint Logic Programming Systems and Logic-based Methods in Programming Environments - CICLOPS-WLPE 2010, Edinburgh, Scotland, United Kingdom, 15/07/10 - 15/07/10.

DNF sampling for ProbLog inference. / Shterionov, Dimitar S.; Kimmig, Angelika; Mantadelis, Theofrastos et al.
2010. Paper presented at 26th International Conference on Logic Programming, ICLP 2010 - Joint Workshop on Implementation of Constraint Logic Programming Systems and Logic-based Methods in Programming Environments - CICLOPS-WLPE 2010, Edinburgh, Scotland, United Kingdom.

Research output: Contribution to conference › Paper › Other research output

TY - CONF

T1 - DNF sampling for ProbLog inference

AU - Shterionov, Dimitar S.

AU - Kimmig, Angelika

AU - Mantadelis, Theofrastos

AU - Janssens, Gerda

PY - 2010

Y1 - 2010

N2 - Inference in probabilistic logic languages such as ProbLog, an extension of Prolog with probabilistic facts, is often based on a reduc- tion to a propositional formula in DNF. Calculating the probability of such a formula involves the disjoint-sum-problem, which is computation- ally 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 probabilis- tic logic systems. The paper also presents an experimental comparison with another sampling based inference method previously introduced for ProbLog.

AB - Inference in probabilistic logic languages such as ProbLog, an extension of Prolog with probabilistic facts, is often based on a reduc- tion to a propositional formula in DNF. Calculating the probability of such a formula involves the disjoint-sum-problem, which is computation- ally 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 probabilis- tic logic systems. The paper also presents an experimental comparison with another sampling based inference method previously introduced for ProbLog.

UR - https://www.scopus.com/pages/publications/84886643653

M3 - Paper

AN - SCOPUS:84886643653

T2 - 26th International Conference on Logic Programming, ICLP 2010 - Joint Workshop on Implementation of Constraint Logic Programming Systems and Logic-based Methods in Programming Environments - CICLOPS-WLPE 2010

Y2 - 15 July 2010 through 15 July 2010

ER -

DNF sampling for ProbLog inference

Abstract

Conference

Other files and links

Fingerprint

Cite this