Welcome to this small info space for my biased but honest view
of the uncertain world of "data mining" ! The sole purpose for
this
space is to hold some facts, opinions and reflections that I collected
on
the way to answer the simple but tough question: Where data mining is
going? There are two interweaving themes running here: the application
and enabling technologies, their today and tomorrow. They're
organized into followed sections:
Frontline: What's happening in the fields
Backyard:What's happening in the labs (mostly applied research labs)
Footprint: Successful or not-so-successful stories of data mining
application from the field
Opinion: Views from veterans and experts
Random Thoughts: A place devoted to my spontaneous thoughts
What's your definition of "data mining" here?
Regardless so many similar or different definitions out there, my
definition of data mining is the range of data driven approaches (with
certain level of automation)which may help human being to develop
insight and/or make better decision. Guided by this definition, I would
like embrace a full range of technologies, whether it is called "data
mining", "machine learning", "artificial intelligence" or those
relevant with buzz words "business intelligence" "computational
intelligence" "predicative analytics" etc. Likewise, I would pay
attention to all walks of life which have already use or anticipate to
use this kind of technologies, be it science discovery, engineering,
business or public interest.
What are you biasing toward?
Given my background and interest, my bias is toward the application of
technology. So this place is lighter on the details of technology, to
the extent of its limitation and potential value in the context of
application. In addition, I would devote much space (hopefully) to the
real
world happening and trends. As another bias: instead of focusing on a
point solution, I would try to form a system perspective to the problem
and solution, in other words, put myself in the shoes of the owner of
the
field application, worrying more about the business impact, utility,
pay off or return of investment. Also, because of the inevitable
"information overloading" problem, I can't see the full picture. So
what records here is limited to what I see and experience, that's just
another source of bias.
How can you be honest as the same time of being biased?
I can afford to be honest because I am not in the game currently which
means I don't have any external force to push me biased.