Data Mining in Semi-Structured Data/ Algorithms for Frequent Sets

November 17, 2003
2:50pm - 4:00pm
Halligan 111

Abstract

The speakers will give two short talks on their work in data mining.

Hiroki Arimura: A brief Introduction to Mining Semi-Structured Data
The talk will discuss data mining in domains such as web pages and XML documents. In these domains the information is not "flat" but is also not well structured - thus the name semi-structured data. The talk will review the problems and algorithmic ideas in this area.

Takeaki Uno: Algorithms for Finding Frequent Sets in Data
Mining association rules prevalent in data is one of the main areas of work in data mining. The major sub task in finding rules is finding sets of "events" which are sufficiently frequent in the data. The talk will discus's a new algorithm and implementation for the problem and its performance on real world and artificial datasets.