Probabilistic logic programming

Probabilistic logic programming
Probabilistic logic programming 2017
An
ILP
workshop
Orléans, France - 7 September 2017
NEW:
Special issue
There will be a special issue on Probabilistic Logic Programming in the
International Journal of Approximate Reasoning (
IJAR
). We welcome submissions of
(improved/extended versions of) papers that were presented at the workshop in
Orléans, as well as new submissions on all topics of the workshop.
For more information, please see the
call for papers
at the IJAR website.
Deadline:
March 7, 2018
EXTENDED to Aptil 7, 2018 (Papers will be sent to reviewers as soon as we receive them).
Publication of the special issue: January 2019 (tentative).
Programme
9:30-10:30
Invited talk
Inference for Probabilistic Logic Programming with Continuous Distributions
. Arjen Hommersom
10:30-11:00
Coffee break
11:00-11:30
Arnaud Nguembang Fadja, Evelina Lamma and Fabrizio Riguzzi.
Deep Probabilistic Logic Programming
11:30-12:00
Anthony Di Franco. Information-gain computation
12:00-14:00
Lunch break
14:00-15:00
Invited talk
PRISM revisited: declarative implementation of a probabilistic programming language using delimited control
. Samer Abdallah
15:00-15:30
Alexander Vandenbroucke and Tom Schrijvers.
From PRISM to ProbLog and Back Again
15:00-15:30
Marco Alberti, Evelina Lamma, Fabrizio Riguzzi and Riccardo Zese.
A Distribution Semantics for non-DL-Safe Probabilistic Hybrid Knowledge Bases
15:30-16:30
Discussion
Proceedings
The proceedings can be found at
CEUR
Overview
Probabilistic logic programming (PLP) approaches have received much attention
in this century. They address the need to reason about relational domains under
uncertainty arising in a variety of application domains, such as bioinformatics,
the semantic web, robotics, and many more. Developments in PLP include new
languages that combine logic programming with probability theory as well as
algorithms that operate over programs in these formalisms.
PLP is part of a wider current interest in probabilistic programming.
By promoting probabilities as explicit programming constructs, inference,
parameter estimation and learning algorithms can be run over programs which
represent highly structured probability spaces. Due to logic programming's strong
theoretical underpinnings, PLP is one of the more disciplined areas of
probabilistic programming. It builds upon and benefits from the large body of
existing work in logic programming, both in semantics and implementation, but
also presents new challenges to the field. PLP reasoning often requires the
evaluation of large number of possible states before any answers can be produced
thus breaking the sequential search model of traditional logic programs.
While PLP has already contributed a number of formalisms, systems and well
understood and established results in: parameter estimation, tabling, marginal
probabilities and Bayesian learning, many questions remain open in this exciting,
expanding field in the intersection of AI, machine learning and statistics.
This workshop aims to bring together researchers in all aspects of probabilistic
logic programming, including theoretical work, system implementations and
applications. Interactions between theoretical and applied minded researchers
are encouraged. The presence of this workshop at ILP is intended to
encourage collaboration with researchers from the field of Inductive Logic
Programming.
Venue
The workshop will take place at
LIFO, Laboratoire d'Informatique Fondamentale d'Orlèans
. How to reach us is described
here
Registration
The fee for participating in PLP'17 is 40€, which includes coffee breaks and lunch.
The registration for PLP'17 is managed through the ILP registration system. To register, visit
ILP registration
page.
Registration: from
1 May 2017
to
6 September 2017
Invited talks
Arjen Hommersom (Open University, The Netherlands)
Samer Abdallah (Jukedeck Ltd)
Arjen Hommersom: Inference for Probabilistic Logic Programming with Continuous Distributions
Probabilistic logics combine the expressive power of logic with the ability to
reason with uncertainty. Given the discrete nature of logical languages, it is
natural to use these languages for modelling discrete distributions. However,
for many practical applications, both discrete and continuous distributions
are required. In this talk, I will discuss several approaches for
dealing with continuous distributions as part of a PLP language. Furthermore,
I will discuss in more detail our recent PLP approach for dealing with
continuous distributions by means of probability intervals. In particular, the
Interative Hybrid Probabilistic Model Counting (IHPMC) algorithm will be
discussed, which enables approximating a large class of hybrid problems with a
bounded error. It has been shown that the current implementation can outperform
current sampling implementation, in particular when the programs contain
sufficient logical structure.
Samer Abdallah: PRISM revisited: declarative implementation of a probabilistic programming language using delimited control
PRISM is a probabilistic programming language based on Prolog augmented with primitives to
represent probabilistic choice. PRISM is implemented using a combination of low
level support from a modified version of B-Prolog, source level program transformation, and
libraries for probabilistic inference and learning implemented in the imperative language C.
More recently, developers of probabilistic languages working in the functional programming
paradigm have taken the approach of
embedding
probabilistic primitives into an existing
language, with little or no modification to the host language, primarily by using
continuations
captured continuations represent pieces of the probabilistic program which can be manipulated to
achieve a great variety of computational effects.
In this talk, I will describe an approach based on delimited control operators recently introduced
into SWI Prolog. These are used to create a system of nested
effect handlers
which together
implement a core functionality of PRISM -the building of explanation graphs- entirely in Prolog
and using an order of magnitude less code. In addition, other declarative programming tools, such
as constraint logic programming, are used to implement several tools for inference, such as the
inside-outside and EM algorithms, lazy best-first explanation search, and MCMC samplers.
By embedding the functionality of PRISM into SWI Prolog, users gain access to its rich libraries and
development environment. By expressing the functionality of PRISM in a relatively small amount of
pure, high-level Prolog, this implementation will hopefully facilitate further experimentation with the
mechanisms of probabilistic logic programming and extensions to new modelling features.
Dates
Papers due:
Sun,
11th 18th
21st June 2017
(EXTENDED)
Notification to authors:
Tue, 11th July 2017
Camera ready version due:
Tue, 25th July 2017
Registration deadline:
Wed, 6th September 2017
Workshop date:
Thu, 7th September 2017
(the deadline for all dates is 23:59 BST)
Paper submissions
Full papers: 6-12 pages, short communications 2-5 pages.
Submissions site:
easychair
Call for papers:
txt
Programme committee
Chairs
Christian Theil Have
(Copenhagen University, Denmark)
Riccardo Zese
(University of Ferrara, Italy)
Programme committee
Christian Theil Have (Copenhagen University, Denmark) [co-chair]
Riccardo Zese (University of Ferrara, Italy) [co-chair]
Samer Abdallah (Jukedeck Ltd)
Elena Bellodi (University of Ferrara, Italy)
Fabio Cozman (University of Sao Paulo, Brasil)
Yoshitaka Kameya (Meijo University, Japan)
Matthias Nickles (NUI Galway, Ireland)
Aline Paes (Universidade Federal Fluminense, Brazil)
Taisuke Sato (NII/SONAR, Japan)
Herbert Wiklicky (Imperial College London, UK)
Theresa Swift (CENTRIA, Portugal)
Senior Committee
Nicos Angelopoulos (Sanger Institute, UK)
Vitor Santos Costa (Universidade do Porto, Portugal)
James Cussens (University of York, UK)
Arjen Hommersom (Open University, The Netherlands)
Angelika Kimmig (KU Leuven, Belgium)
Evelina Lamma (University of Ferrara, Italy)
David Poole (University of British Columbia, Canada)
Luc De Raedt (KU Leuven, Belgium)
Fabrizio Riguzzi (University of Ferrara, Italy)
Alessandra Russo (Imperial College, UK)
Joost Vennekens (KU Leuven, Belgium)
Last modified: Fri 6 October 2017