2112CIT CONCEPTUAL FOUNDATIONS OF ARTIFICIAL INTELLIGENCE

COURSE OUTLINE

Course catalogue no:  2112CIT
Course title: Conceptual Foundations of Artificial Intelligence
Field of Education Code: 020119
Program/s: 1042 Bachelor of Information Technology
1043 Bachelor of Information Technology with Advanced Studies
1045 Bachelor of Science/Bachelor of Information Technology
1151 Bachelor of Engineering in Microelectronic Engineering/Bachelor of Information Technology
School: School of Computing and Information Technology
Faculty: Faculty of Engineering and Information Technology
Status of Course within program/s or academic plan/s: Core second year Artificial Intelligence major
Credit point value: 10CP
Prerequisites:  1103CIT Programming 1
1109CIT Introduction to Artificial Intelligence
Year and semester: Semester 1, 2005
Course Convenor: Dr Terry Dartnall
Room 1.27, N44
(07) 387 55020
T.Dartnall@griffith.edu.au
Teaching team members: Dr Terry Dartnall
Date course outline was last modified: 2005, Semester 1

ORGANISATION AND TEACHING METHODS

Two hours of lectures and one 2 hour laboratory per week.

OBJECTIVES

This course builds on 1109CIT Introduction to Artificial Intelligence by further examining the issues, problems and approaches that were introduced in first year and by introducing new ones.  It emphasises implementation and teaches students to implement methods and techniques that were introduced in 1109CIT.  It introduces one of the procedural programming languages used in AI applications and provides hands on experience in neural computation.

This is a required course in the Artificial Intelligence major.

INTERRELATIONSHIP OF THE COURSE WITH OTHER COURSES AND THE PROGRAM

This course extends the theory that was taught in 1109CIT Introduction to Artificial Intelligence and shows how to implement this theory in modular programs that integrate AI methods such as search, knowledge representation and natural language processing.  This integrates theory and practice.  It also shows how intelligence requires the integration of modular abilities.  This in turn shows how abilities such as reasoning, knowledge representation, learning and vision that are addressed in other courses in the major can be integrated in a single intelligent system.

BRIEF DESCRIPTION

This course builds on 1109CIT by further examining problems and approaches that were introduced in first year and by introducing new ones. It introduces one of the procedural languages used in AI applications and provides hands-on experience in neural computation.

CONTENT

The course will introduce Pop-11 or Lisp, and will cover a selection of the following topics:
 
  •  search
  •  planning
  •  knowledge representation
  •  natural language processing
  •  connectionism (neural computation)
  •  artificial life
  •  cognitive modelling in AI
  •  philosophical issues in AI

    • GENERIC SKILLS DEVELOPMENT

    The course teaches problem solving in the writing and integration of medium to large scale modular programs.  It emphasises the importance of clarity and accessibility.  Assignments must be clearly annotated, so that the code is transparent and accessible.  Assignments must be accompanied by an introduction that clearly outlines the modular structure of the program.  Above all, programs must be usable.  The major assignment is written by pairs of students.  This encourages them to work as members of a team.

    FLEXIBLE LEARNING

    The teaching material is available online.  Most of it is available in teach and help files that I have written for the course in the Poplog Artificial Intelligence teaching and programming environment.  These files contain detailed information and well as executable code, so that students can study the theory and see the programming code being developed at the same time.  To a large extent this enables students to proceed at their own pace.  Students may make out a case for being allowed to do the major assignment by themselves rather than in pairs.

    RATIONALE FOR CONTENT

    The balance of two hours of lectures and one 2 hour laboratory per week reflects the balance of theory and application in the course.

    ASSESSMENT

    Assessment Item Worth Focus Due Generic Skill(s) Addressed
    Programming Assignment 1 15% Individual  End week 5 Problem solving
    Programming Assignment 2 35% Group End week 9 Problem solving.  Working as a team member
    Final Examination 50% Individual End of semester Analysis and critical evaluation.  Problem solving

    Students must get at least half the marks available in the final examination in order to achieve a grade of Pass or better in the course.

    RATIONALE FOR ASSESSMENT

    The first assignment assesses the student's grasp of some basic AI programming techniques. The second assignment assesses the student's ability to integrate more sophisticated methods into a single program. Assignments are written in laboratories under the supervision of the teaching team.

    Students must get at least half the marks available in the final examination in order to achieve a grade of Pass or better in the course.  This ensures that they understand the theory on which their assignments are based.

    TEXTS AND SUPPORTING MATERIAL

    The  text book is:

    Sharpes l, M., et. al. (eds) Computers and Thought: A Practical Introduction to Artificial Intelligence, MIT/Bradford, 1989.

    There is an online version of this book.  It is broken up into a large number of nodes which can make it frustrating to use - but it's cheaper than buying the textbook.

    The book is a general introduction to AI but is designed to be used in conjunction with the AI environment Poplog.  Poplog is a huge environment that includes the programming languages Pop11 and Prolog (hence "Poplog"), as well as Common Lisp.  It contains thousands of teaching and other files, and the editors ved and xved.  Because Poplog files contain executable code, i.e. programs that run when you compile them, there is no point in installing them on the Web.  The easiest way to orient yourself in Poplog (in which you can easily get lost) is to go to HELP LOCAL, which is a directory of the teaching files I've installed on the system.  The index at the beginning of HELP LOCAL directs you to the 2112CIT Conceptual Foundations section .  The first teaching file in this section is HELP COMMANDS, which is a mismash of ved, xved and Unix commands.  The ved commands are something of a throwback and you don't really need them, but they're handy if you can't use windows on your modem.

    One way to get into Poplog is to type "xved" at the Unix prompt.  This will throw up a window.  The right-hand enter button will take you to the command line at the top of the file.  Type "help local" <return> on the command line and a window will be thrown up.  <enter> im <return> gives you the immediate mode window.  <enter> ved filename.p <return> gives you a file-mode window.  Or you can xved directly into a file from the Unix prompt, e.g. "xved filename.p".  Play around:  you can have as many windows as you like.

    Neural computation will be taught using the Brainwave neural computation environment developed at the University of Queensland. The BrainWave Handbook tells you all about BrainWave. BrainWave itself can be started by clicking here.  To quit out of it, pull down the file menu and click on "exit".

    Timetable

    A course aims and approximate weekly syllabus is  also available.


    [back]School of Information and Communication Technology
    [back]Griffith University
    Author: Terry Dartnall.
    Last modified: 16/3/2005. If you have any comments please contact T.Dartnall@griffith.edu.au