=Paper=
{{Paper
|id=Vol-1661/invited-02
|storemode=property
|title=None
|pdfUrl=https://ceur-ws.org/Vol-1661/invited-02.pdf
|volume=Vol-1661
}}
==None==
<pdf width="1500px">https://ceur-ws.org/Vol-1661/invited-02.pdf</pdf>
<pre>
Probabilistic Inductive Logic Programming
   Based on Answer Set Programming

                              Matthias Nickles

INSIGHT Centre for Data Analytics & Discipline of Information Technology
                National University of Ireland, Galway



  Abstract. Answer Set Programming (ASP) is a form of declarative pro-
  gramming based on the concept of so-called stable models of programs,
  with roots in logic programming and nonmonotonic reasoning. ASP has
  emerged as a fully declarative programming paradigm which provides
  significant advantages in areas such as search and optimization prob-
  lem solving, common sense knowledge representation, and modeling of
  nondeterminism. In my talk, I will describe how ASP can be used as a
  basis for expressive probabilistic inductive logic programming, and the
  features (and challenges) of this direction. After introducing ASP and
  providing an overview of existing approaches to probabilistic declarative
  programming based on stable model semantics, I will present a recent
  framework for probabilistic inductive ASP which provides a high level
  of expressiveness (including the option to use first-order formulas with
  probabilities) in combination with a high degree of adaptability to a vari-
  ety of tasks. I will discuss algorithms for inference and machine learning
  in this framework and their respective performance characteristics, and
  present possible applications of our framework. The last part of my talk
  will outline directions for future research in this area.

</pre>