|The idea for this repository originated during the
1994 AAAI Symposium on
Improving Instruction of Introductory AI held in New Orleans in November 1994. The
symposium was organized by Marti Hearst (Chair), Haym Hirsh, Dan Huttenlocher, Nils
Nilsson, and Bonnie Webber. One of the main goals of the symposium was to provide a forum
for the exchange of ideas/experiences among instructors teaching the course and facilitate
discussions on related issues. Some of the main issues addressed by the symposium
participants were: (a) themes used to structure the AI course, (b) course goals and
outcomes, (c) the role of programming in the AI course, and (d) course content and its
place in the undergraduate CS curriculum.
A similar forum took place at the 1995 NSF Workshop on Providing and Integrating Educational Resources for Faculty Teaching Artificial Intelligence, and its subsequent Reflection Day held in Philadelphia in June 1995, and February 1996, respectively. This workshop was organized by Robert Aiken, Daniel Hardt, Giorgio Ingargiola, Deepak Kumar, Sarah Rebecca Thomas, Roger Webster, and Judith Wilson. During this workshop additional ideas were discussed and materials were made available for the repository.
After this workshop Bill Manaris and Ingrid Russell (with help and feedback from Mike Hamilton) developed a prototype which played a major role in establishing the design/structure, and contents of this repository.
|During the above meetings, it was generally agreed
that preparing material for AI courses is a major challenge. This is mostly due to the
diversity of the topics that might be covered. This difficulty is more pronounced at
institutions that do not have a strong AI research core to provide the discussion support
needed by those teaching the course. Additionally, the diversity of student backgrounds,
as they enter the course, raises an additional challenge in preparing materials for and
deciding which resources to use in the course. Finally, in addition to determining what to
cover in the course and how to present the material so that topics connect together as a
unified whole, a major challenge faces instructors of AI courses. This challenge is how
to locate available educational resources and material for the course.
Within the last few years, due to the rapid development of the World-Wide-Web (WWW), a wealth of material has become available that could be used in AI pedagogy. Although numerous AI resources are available, a large percentage of them are commercial and production-environment oriented and thus not necessarily appropriate for or desirable in AI pedagogy.
|This repository is a centralized distribution
point which (a) focuses on WWW educational resources and material to be used in
undergraduate AI pedagogy, (b) identifies and organizes such resources by topic, and (c)
provides a wide selection of resources applicable to various student expertise levels and
budgetary considerations/constraints. Specifically, it contains:
|While researching the currently available
resources related to AI, we discovered that the majority of them are geared towards
production environments as opposed to education. Since it would appear that it would take
much longer for students (and the instructor) to get acquainted with the former, their
adoption may be undesirable for a regular one-semester/quarter AI course. However, for
graduate-level courses where a more advanced and focused approach is desired, it might be
appropriate to concentrate more on production-oriented systems. Therefore, the repository
attempts to provide a balanced mix of resources in terms of required hardware/software
platforms, acquisition cost, system complexity, and specific AI topics.
In terms of platform, we include systems that are available for as many platforms as possible. We believe that, the more platforms a system has been ported to, the stronger the indication that (a) the system is mature/stable enough in terms of functionality, (b) the system is supported by its developers, and (c) more educators and students could benefit from its inclusion in the repository. In some cases, however, we may deviate from this heuristic, due to exceptional reasons, such as uniqueness in a topic area, system development philosophy, and price range.
In terms of price, we include a balanced mix between freeware/shareware and commercial systems. Clearly, the fact that a system is free may have nothing to do with its quality or support commitment from its developer(s). For commercial systems, we decided to include systems costing less than $5,000, since we believe that very few schools might consider investing more than this amount in a specialized educational resource for a single undergraduate course.
In terms of complexity, we provide a wide spectrum of systems, ranging from very simple ones that could be used for a single assignment, to more advanced ones that could be used for a sequence of assignments or a subsequent special topics course.
In terms of topic and classification of these resources, we use the following taxonomy: Expert Systems, Fuzzy Logic, Knowledge Representation, Logic and Reasoning, Natural Language Processing, Neural Networks, Robotics, Search and Game Playing, Vision, and Other. This taxonomy is not necessarily the only possible one, since there exist many opinions on which general topics constitute AI, and what are the names and scopes of these topics. Nevertheless, the above taxonomy is consistent what the majority of available textbooks and course syllabi that we examined.
|This site will continue to grow as long as
educators continue to share their work. To send your contributions or suggestions, please
use the automated submission forms (resources, tools and environments, class
syllabi, texbooks, suggestions).
|The above has been adapted from
visitors since April 8, 1999 (counter provided by LinkExchange)