DM: Introductions

[Miguel Feldens]

Hi

As has been suggested, I would like to introduce myself to the people
subscribing this list.

I am associate researcher at Universidade Federal do Rio Grande do Sul
(Porto Alegre, Brazil), and my main interest is data mining. My 
master's
dissertation has been on "The Engineering of Knowledge Discovery in
Databases: a case study in a health care domain", including the 
design,
development and validadion of a data mining prototype, applied to 
health
care data. Right now, we are working on this system, in order to 
transfer
data mining technology to government's health care system, aiming to
discover useful information from hospital administration standpoint, 
as
well as detecting errors/frauds.

The approaches that have been explored in our work so far are 
association
rules and visualization techniques, in order to make discovered 
knowledge
understandable for a wider range of users (in our case, medical 
doctors,
hospital administrators, as well as DBAs). Our research prototype is
currently capable to discover all association rules given certain
user-defined constrains, and present to the user as a list of rules, 
as
knowledge graphs, and a "browseable" hierarchycal structure 
containing the
rules.

We are also interested in issues as data mining components, agents and
requirements for the automation of KDD process.

Thank you all for the opportunity to exchange experience!

Miguel Feldens
Universidade Federal do Rio Grande do Sul
Brazil
feldens@inf.ufrgs.br
http://www.inf.ufrgs.br/~feldens/