Andrew Trotman, PhD Student, Department of Computer Science
Title: "Genetic Learning for Information Retrieval"
Online literature searching is something we all know and hate. It takes to long, search results are notoriously bad, and its hard to find what you're looking for. Should suitable training sets be available, genetic learning techniques (GAs and GPs) could be used to improve this. Such learning sets do exist. Focusing on document ranking (the precision metric), an application of genetic learning to information retrieval will be discussed and results showing improved rankings presented.
Slides may be viewed here