KLAP Seminar Home Page
Spring 2004
This is a weekly seminar which discusses issues related to
the areas of logic and constraint programming, knowledge representation,
and parallel processing.
It is organized by members of the
Knowledge representation, Logic, and Advanced Programming Laboratory (KLAP).
The seminar is open to everyone.
Time: Monday, 12:30 pm
Location: SH 124
Upcoming Seminar
There will be no more talks this semester.
Thanks to all of our speakers and participants!
The seminar will resume in Fall 2004.
Schedule of talks
Date | Speaker | Title
|
February 18
| Omar El Khatib
| ASP-PROLOG: A System for Reasoning about Answer Set Programs in Prolog
|
February 25
| Islam Elkabani
| Smodels with CLP and its Applications: A Simple and
Effective Approach to Aggregates in ASP
(Abstract)
|
March 1
| Xiaofeng Xiao
| Parallel Algorithm for Programming in CLP(SET)
(Abstract)
|
March 8
| Ian Strascina
| "MINERVA - A Dynamic Logic Programming Agent Architecture"
by Leite, Alferes, Pereira
(
talk slides,
paper,
paper on
LUPS)
|
March 15
| Chongbing Liu
| Inductive Logic Programming: Basic Approaches
(talk slides)
|
March 29
| Brian Cloteaux
| Dynamically Finding Least Common Ancestors in DAGs
|
April 5
| Hung Viet Le
| Induction of Logic Programs: FOIL and Related Systems
by J. R. Quinlan and R. M. Cameron-Jones
(paper)
|
April 12
| Emad Saad
| Hybrid Probabilistic Programs
|
April 19
| Brian Palmer
| Translation and Navigation in Digital Mathematics
|
April 26
| Tu Phan
| Sensing and NonDeterministic Actions in Domains with Multi-Valued
Fluents
by Tran Cao Son and Phan Huy Tu
(paper)
|
Previous Seminars
April 26
Tu Phan
spoke on
Sensing and NonDeterministic Actions in Domains with Multi-Valued
Fluents
by Tran Cao Son and Phan Huy Tu
Abstract
We develope a high-level action description language, called AM_K,
with a transition function based semantics that is capable of
representing and reasoning with (i) non-deterministic
actions; (ii) sensing actions; and (ii)
multi-valued fluents. We demonstrate the use of AM_K through
examples. To use the language in a logic-programming based planner,
we develop an approximation of the entailment relationship between
action theories and queries in AM_K for a
subclass of AM_K action theories.
April 19
Brian Palmer
spoke on
Translation and Navigation in Digital Mathematics
Abstract
This talk presents an electronic mathematics translation and
navigation service I'm developing to improve mathematics accessibility
for the blind and visually impaired. I'll discuss the various
mathematical formats in use today and their relationship to one
another, and discuss methods for classifying strctures in documents
for navigation purposes.
April 12
Emad Saad
spoke on
Hybrid Probabilistic Programs
Abstract
In this talk I will present Dekhtyar and Subrahmanian approach for
probabilistic logic programming; Hybrid Probabilistic Programs. Quoting
Dekhtyar and Subrahmanian, hybrid probabilistic logic programs is a
variation of the probabilistic annotated logic programming approach,
which allows the user to explicitly encode the available knowledge
about the dependency among the events in the program. In this talk the
syntax and semantics of hybrid probabilistic programs will be presented.
April 5
Hung Viet Le
spoke on
Induction of Logic Programs: FOIL and Related Systems
by J. R. Quinlan and R. M. Cameron-Jones
Abstract
FOIL is a first-order learning system that uses information
in a collection of relations to construct theories expressed in a
dialect
of Prolog. This paper provides an overview of the principal ideas and
methods used in the current version of the system, including two recent
additions. We present examples of tasks tackled by FOIL and of systems
that adapt and extend its approach.
March 29
Brian Cloteaux
spoke on
Dynamically Finding Least Common Ancestors in DAGs
Abstract
A common problem in the parallelization of logic programs is to
dynamically find the least common ancestor for two nodes in a computation
tree. This has been a well understood problem for a number of years.
But if we want to generalize this problem to dynamically finding the least
common ancestor in a directed acyclic graph (DAG), we then find that there
is almost no published research in this area. We will discuss a pair of
recent results which help us to understand why this problem is difficult,
and show how a dynamic algorithm can be created to solve it.
March 15
Chongbing Liu
spoke on
Inductive Logic Programming: Basic Approaches
Abstract
This talk gives a very general introduction to ILP
and presents the basic approaches of ILP in certain
amount of details. In the first part, the relation
between ILP and Machine Learning as well as Logic
Programming is illustrated and the problem specification
of ILP is given. Then the structure, properties and
operations of ILP search space are discussed. In the
second part, basic top-down and bottom-up approaches
are presented by corresponding algorithms. More attention
is paid on the bottom-up approaches, including those
using Relative Least General Generalization techniques
and Inverse Resolution. For the purpose of demonstration,
a very simple ILP problem is solved using Progol system
in the third part. Finally, the future directions of ILP
are discussed, intending to believe that more research
effort should put on new search approach invention and
parallel implementation.
March 8
Ian Strascina
spoke on
MINERVA - A Dynamic Logic Programming Agent Architecture
Abstract
The agent paradigm, commonly implemented by means of imperative
languages mainly for reasons of efficiency, has recently increased its
influence in the research and development of computational logic based
systems. Since efficiency is not always the crucial issue, but clear
specification and correctness is, Logic Programming and Non-monotonic
Reasoning (LPNMR) have been brought back into the spotlight. To this
accrues the recent significant improvements in the efficiency of Logic
Programming implementations for Non-monotonic Reasoning.
This paper presents an overall description of MINERVA, an agent
architecture and system designed with the intention of providing a common
agent framework based on the unique strengths of Logic Programming, to
allow for the combination of several non-monotonic knowledge
representation and reasoning mechanisms developed in recent years.
March 1
Xiaofeng Xiao
spoke on
Parallel Algorithm for Programming in CLP(SET)
Abstract
We present the resulting constraint algorithms which are embedded in a
Constraint Logic Programming (CLP) language which provides finite sets -- along
with basic set-theoretic operations -- as first-class objects of the language.
We also provide the parallel algorithm for programming in CLP(SET) which will
improve the performance of programming in CLP(SET) greatly. Analysis of
multiple
choices, like using MPI or PTHREAD, in the programming will also be presented.
Further works include improving the cooperation between robots and designing
more efficiency ground-base missile defense system using the parallel algorithm
for programming in CLP(SET).
February 25
Islam Elkabani
spoke on
Smodels with CLP and its Applications: A Simple and
Effective Approach to Aggregates in ASP
Abstract
In this work we propose a semantically well-founded extension of
Answer Set Programming (ASP) with aggregates, which relies
on the integration between state-of-the-art answer set solvers and
constraint logic programming systems. The resulting system is efficient,
flexible, extensible, and supports form of aggregation more general than
those previously proposed in the literature. The system is developed as
an instance of a general framework for the embedding of arbitrary
constraint theories within ASP.
February 18
Omar El Khatib
spoke on
ASP-PROLOG: A System for Reasoning about Answer Set Programs in Prolog
Abstract
We present a system, called ASP-PROlOG, which provides a tight
and well-defined integration of Prolog and Answer Set Programming (ASP).
The combined system enhances the expressive power of ASP, allowing
programmers to write programs that reason about ASP modules, that can be
dynamically updated, and about collection of stable models. This feature
is vital in a number of application domains (e.g., planning, scheduling,
diagnosis, optimizations). We describe the design of ASP-PROLOG along with
its implementation, realized using CIAO Prolog and Smodels.
This page is maintained by
Inna Pivkina