Module Title

Logic Programming and Artificial Intelligence A and B

Module Code

COM530J1

Module Level

D

Credit Points

20

Semester

1

Module Status

Optional

Location

Jordanstown

Prerequisite(s)

MT216A or MT216B

Corequisite(s)

None

Module Coordinator

Professor I Düntsch

Teaching Staff Responsible

Professor I Düntsch, ISE, Jordanstown

Dr H Wang, ISE, Jordanstown

Contact Hours

Lectures

48 hours

Tutorials

4 hours

Practicals

24 hours

Academic Topic

Private Study

72 hours

Computing

Rationale

This module deals with the scope and limits of systems which exhibit some kind of intelligent behaviour. At the core of these are the formal foundations of automated reasoning and numerical as well as non-numerical methods to handle uncertainty. Prolog programming will implement some of these methods. The A version is taken by honours candidates and the B version by ordinary degree students.

Learning Outcomes

Upon the successful completion of this module a student should be able to:

(i)

Demonstrate familiarity with the possibilities and limits of quasi - intelligent systems., formal foundations of knowledge representation, organisation, manipulation and uncertainty handling,

(ii)

Exemplify the parts of Gigerenzer's data model and sources of uncertainty in the modelling process.

(iii)

Describe and apply techniques of problems solving.

(iv)

Exemplify the parts of a formal system and do simple proofs and derivations.

(v)

Explain Chomsky's hierarchy of generative grammars.

(vi)

Demonstrate a good knowledge of the syntactic and semantic aspects of propositional logics, and be able to prove simple theorems.

(vii)

Discuss fundamental principles of inductive reasoning such as Occam's razor, the principle of indifference, and the maximum entropy principle.

(viii)

Demonstrate a good knowledge of techniques of inductive reasoning such as e.g. the Bayesian approach, and discuss its strength and weaknesses.

(ix)

Demonstrate a good knowledge of techniques of rule based reasoning such as e.g. the rough set model or qualitative scaling, and discuss its strength and weaknesses.

(x)

Demonstrate a mastery of basic Prolog programming.

Content

1

Scope and limits of artifacts exhibiting some form of intelligent behaviour

2

Introduction to data modelling

3

Problem solving strategies

4

Formal languages and systems

5

Grammars and logics

6

Principles of inductive reasoning and uncertainty

7

Numerical reasoning methods

8

Symbolic reasoning methods

9

Elements of Prolog programming

Learning and Teaching Methods

A combination of lectures, tutorials, lab sessions, and individual counselling. At the beginning of each section the students receive a list of objectives for the current area, keywords and key symbols (via the WWW). Lecture notes are distributed electronically in PDF format.

Assessment:

Coursework Assignments 25% (A), 40% (B)

Written Examination 75% (A), 60% (B)

Coursework 1:

An open book, open ended class test covering learning outcomes (i) - (iii)

Coursework 2 (COM530J1A only):

An open book, open ended class test covering learning outcomes (iv) - (vi)

Coursework 3:

An open book, open ended class test covering learning outcomes (vii) - (ix)

Coursework 4:

Graded assignments in Prolog covering learning outcome (x)

Examination:

The exam paper consists of seven questions in three parts: Part A consists of a problem with 20 multiple choice questions (15 for COM530J1B) covering all learning outcomes. Part B consists of three problems (two for COM530J1B) which cover a choice of material. Part C covers Prolog programming, and consists of three problems (two for COM530J1B). Part A is compulsory, and at least one problem from sections B and C must be attempted. For full marks, a student has to solve four problems (three for COM530J1B) altogether from Sections B and C.

Overall Mapping of learning outcomes for COM530J1 on BSc Hons MSC

(i)

(ii)

(iii)

(iv)

(v)

(vi)

(vii)

(viii)

(ix)

(x)

COM530J1Aex

x

x

x

x

x

x

x

x

x

x

COM530J1Acw

x

x

x

x

x

x

x

x

x

x

Overall Mapping of learning outcomes for COM530J1 on BSc MSC

(i)

(ii)

(iii)

(iv)

(v)

(vi)

(vii)

(viii)

(ix)

(x)

COM530J1Bex

x

x

x

x

x

x

x

COM530J1Bcw

x

x

x

x

x

x

x

Reading List.

Recommended

Ivan Bratko "Prolog - Programming for Artificial Intelligence"

I Düntsch and G Gediga, "Sets, relations, functions"

S Russell and P Norvig, "Artificial Intelligence"

T Munakata, "Fundamentals of the New Artificial Intelligence"

Required

I Düntsch and G Gediga, "An introduction to the theory of building intelligent artifacts". Distributed electronically.