Loglinear models for first-order probabilistic reasoning

J Cussens

Loglinear models for first-order probabilistic reasoning

Computer Science

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

Recent work on loglinear models in probabilistic constraint logic programming is applied to first-order probabilistic reasoning. Probabilities are defined directly on the proofs of atomic formulae, and by marginalisation on the atomic formulae themselves. We use Stochastic Logic Programs (SLPs) composed of labelled and unlabelled definite clauses to define the proof probabilities. We have a conservative extension of first-order reasoning, so that, far example, there is a one-one mapping between logical and random variables. We show how, in this framework, Inductive Logic Programming (ILP) can be used to induce the features of a loglinear model from data. We also compare the presented framework with other approaches to first-order probabilistic reasoning.

Original language	English
Title of host publication	UNCERTAINTY IN ARTIFICIAL INTELLIGENCE, PROCEEDINGS
Editors	KB Laskey, H Prade
Place of Publication	SAN FRANCISCO
Publisher	MORGAN KAUFMANN PUB INC
Pages	126-133
Number of pages	8
ISBN (Print)	1-55860-614-9
Publication status	Published - 1999
Event	15th Conference on Uncertainty in Artificial Intelligence - STOCKHOLM Duration: 30 Jul 1999 → 1 Aug 1999

Conference

Conference	15th Conference on Uncertainty in Artificial Intelligence
City	STOCKHOLM
Period	30/07/99 → 1/08/99

Keywords

loglinear models
constraint logic programming
inductive logic programming

Cite this

@inproceedings{763739cb70ad4bdd9875d976db63e261,

title = "Loglinear models for first-order probabilistic reasoning",

abstract = "Recent work on loglinear models in probabilistic constraint logic programming is applied to first-order probabilistic reasoning. Probabilities are defined directly on the proofs of atomic formulae, and by marginalisation on the atomic formulae themselves. We use Stochastic Logic Programs (SLPs) composed of labelled and unlabelled definite clauses to define the proof probabilities. We have a conservative extension of first-order reasoning, so that, far example, there is a one-one mapping between logical and random variables. We show how, in this framework, Inductive Logic Programming (ILP) can be used to induce the features of a loglinear model from data. We also compare the presented framework with other approaches to first-order probabilistic reasoning.",

keywords = "loglinear models, constraint logic programming, inductive logic programming",

author = "J Cussens",

year = "1999",

language = "English",

isbn = "1-55860-614-9",

pages = "126--133",

editor = "KB Laskey and H Prade",

booktitle = "UNCERTAINTY IN ARTIFICIAL INTELLIGENCE, PROCEEDINGS",

publisher = "MORGAN KAUFMANN PUB INC",

note = "15th Conference on Uncertainty in Artificial Intelligence ; Conference date: 30-07-1999 Through 01-08-1999",

}

TY - GEN

T1 - Loglinear models for first-order probabilistic reasoning

AU - Cussens, J

PY - 1999

Y1 - 1999

N2 - Recent work on loglinear models in probabilistic constraint logic programming is applied to first-order probabilistic reasoning. Probabilities are defined directly on the proofs of atomic formulae, and by marginalisation on the atomic formulae themselves. We use Stochastic Logic Programs (SLPs) composed of labelled and unlabelled definite clauses to define the proof probabilities. We have a conservative extension of first-order reasoning, so that, far example, there is a one-one mapping between logical and random variables. We show how, in this framework, Inductive Logic Programming (ILP) can be used to induce the features of a loglinear model from data. We also compare the presented framework with other approaches to first-order probabilistic reasoning.

AB - Recent work on loglinear models in probabilistic constraint logic programming is applied to first-order probabilistic reasoning. Probabilities are defined directly on the proofs of atomic formulae, and by marginalisation on the atomic formulae themselves. We use Stochastic Logic Programs (SLPs) composed of labelled and unlabelled definite clauses to define the proof probabilities. We have a conservative extension of first-order reasoning, so that, far example, there is a one-one mapping between logical and random variables. We show how, in this framework, Inductive Logic Programming (ILP) can be used to induce the features of a loglinear model from data. We also compare the presented framework with other approaches to first-order probabilistic reasoning.

KW - loglinear models

KW - constraint logic programming

KW - inductive logic programming

M3 - Conference contribution

SN - 1-55860-614-9

SP - 126

EP - 133

BT - UNCERTAINTY IN ARTIFICIAL INTELLIGENCE, PROCEEDINGS

A2 - Laskey, KB

A2 - Prade, H

PB - MORGAN KAUFMANN PUB INC

CY - SAN FRANCISCO

T2 - 15th Conference on Uncertainty in Artificial Intelligence

Y2 - 30 July 1999 through 1 August 1999

ER -