The 22nd Australasian Joint Conference on Artificial Intelligence

Keynote Speakers

Prof. Mark Bedau
Reed College, Oregon, USA

Title: The second creation: the scientific and social implications of making new forms of life in the laboratory.

Abstract: A protocell is a microscopic minimal chemical system that assembles itself, grows and reproduces similar daughter protocell. The process of natural selection operating on a population of protocells could adapt and improve their ability to survive and reproduce. This talk selectively surveys the state of the art in protocell research and development, and sketches the social and ethical implications of making new forms of life in the laboratory.

Prof. Andries P. Engelbrecht
University of Pretoria, Pretoria, South Africa

Title: CIlib: A Component-based Framework for Plug-and-Simulate Computational Intelligence Systems

Abstract: Research in Computational Intelligence (CI) has produced a huge collection of algorithms, grouped into the main CI paradigms. These CI algorithms are increasingly being used to create hybrid intelligent systems, where different algorithms from different CI paradigms are combined to form a new model. Implementation of such CI systems requires that the underlying CI algorithms be implemented. While most CI systems have specialized implementations focusing only on specific CI algorithms, the development process usually requires different variations of a CI technique to be implemented and tested in order to find the best combination of CI algorithms for the intelligent system. This process usually demands a re-implementation of existing algorithms, and sometimes even rewrites of the entire system skeleton. In addition to the CI components of complex CI systems, a communications protocal for information (or state) exchange among CI components needs to be defined and implemented. Here it may also become necessary to implement and test different communications protocols. When a final system has been produced, this system has to be thoroughly evaluated and benchmarked against other models. The development and evaluation of a complex CI system can then become a tedious and time consuming process. Furthermore, re-implementation of existing CI algorithms may lead to code bugs and wastes time. Trying to implement a new generic CI system framework for each new research study can become a nightmare.

This presentation will introduce a new, opensource component-based framework which

provides a generic framework to implement any CI algorithm, or variation of that algorithm,
facilitates the process of implementing a generic CI system, where any of the CI algorithms (components) can be used within the system,
provides a generic framework for implementing any communications protocol,
allows easy implementation of the problem to be solved,
provides an XML interface to easily glue components together to form the CI system, and
provides a simulator to manage the process of running a specified number of simulations on all specified benchmark problem

The talk will discuss this library, called CIlib, in detail and will show that it provides an environment for plug-and-simulate CI systems, and doing so with minimal development effort.

A/Prof. Eamonn Keogh
University of California - Riverside, California, USA

Title: AI for Understanding Cultural Heritage: Progress and Challenges

Abstract: Artificial intelligence and data mining are broadly applicable tools, yet they thus far have had very limited impact in help us to understand our cultural heritage. In this talk I argue that the time is now ripe for such an effort, driven by both technical advances and an increasing understanding (and willingness to fund!) of the important of cultural heritage.

Prof. Ian H. Witten
University of Waikato, Hamilton, New Zealand

Title: Semantic document processing using Wikipedia as a knowledge base

Abstract: Wikipedia is a goldmine of information; not just for its many readers, but also for the growing community of researchers who recognize it as a resource of exceptional scale and utility. It represents a vast investment of manual effort and judgment: a huge, constantly evolving tapestry of concepts and relations that is being applied to a host of tasks.

This talk will introduce the process of "wikification"; that is, automatically and judiciously augmenting a plain-text document with pertinent hyperlinks to Wikipedia articles -- as though the document were itself a Wikipedia article. This amounts to a new semantic representation of text in terms of the salient concepts it mentions, where "concept" is equated to "Wikipedia article." Wikification is a useful process in itself, adding value to plain text documents. More importantly, it supports new methods of document processing.

I first describe how Wikipedia can be used to determine semantic relatedness, and then introduce a new, high-performance method of wikification that exploits Wikipedia's 60 M internal hyperlinks for relational information and their anchor texts as lexical information, using simple machine learning. I go on to discuss applications to knowledge-based information retrieval, topic indexing, document tagging, and document clustering. Some of these perform at human levels. For example, on CiteULike data, automatically extracted tags are competitive with tag sets assigned by the best human taggers, according to a measure of consistency with other human taggers.

Although this work is based on English it involves no syntactic parsing, and the techniques are largely language independent. The talk will include live demos.

The 22nd Australasian Joint Conference on Artificial Intelligence (AI'09)

The 22^nd Australasian Joint Conference on Artificial Intelligence (AI'09)