Data Mining
In the
age of Information, almost every company tries to collect all available data.
Only few of them could be able to optimize utilization of the data they collected.
Some of them attempted to summarize the data by visualizing or querying them.
Additional with vast amount of data, it is difficult for them to find the
gold that hides in their databases.
Technique,
derived from Computer Science field, called Data Mining is required to extract
the knowledge that buried deep long ago in the databases. The obtained knowledge
could be used as a competitive edge over your competitors by making you having
more perception on your company and, definitely, customers. The better you
know your customers, the more satisfied offer you make to them.
To
be more descriptive, Data Mining is an approach to find
pattern in the database that may be hardly detected
manually. By knowing pattern of the database, you would
be able to predict what will happen in the future, for
example, you will know what your customers will
do next, what your customers will buy then. This knowledge will make you to
do a much better marketing plan to serve your loyal customers.
By the
fact that there are many kinds of problems in business field, we provide various
kinds of Data Mining services to serve those problems as a specific purpose.
They can be briefly described like this.
Market
Basket Analysis (or Association Rule) -- the tool helps us in finding relations
between at least two events that tend to happen together. We usually do this
approach for cross-selling associated products when a customer expresses interest
in or buys item. You can also bring the analysis result to design promotions,
i.e. bundled packs, or to plan shop layout to persuade customers to buy more
in the way they like.
Clustering
(Segmenting) -- in a huge database, there are always various groups of
similar characters contained. The approach will help us in distinguish dissimilar
records apart and, simultaneously, gather similar items together to be segments
of homogeneous characteristic. Marketing nowadays tend to be more customization,
by segmenting customers into groups of homogeneous behaviour would make companies
to serve their customers more accordance to their needs. Of course, the result
you get is Customer Satisfaction
and Customer Loyalty.
Classification
-- this method is employed when you already have well-defined classes of
data (customers). New record will be classified to an appropriate class. You
will know how to interact with or what to offer to new customer walking in
by just knowing few of his/her profile attributes.
Decision
Tree -- the Tree diagram purely derived from your data will make you easily
understand the happening in your database. The methodology will show you a
tree diagram demonstrating what attributes that affect your interested result,
most to least. One of its advantages is that you can implement your campaign
promotion to a selected group without having to do to all of your customers.
This would help you a lot in saving cost. Also, your customers will be less
annoyed with the campaign offered to them irrelevantly being lessened.
Neural
Network -- the technique is reputable in field of study pattern of the
data. By training Neural Network model examples of past records, it will learn
by itself the relations between all of the factors and the result. After being
well-trained, from the lessons it learned the model could be used to predict
what will happen in the future. Some business applications of neural network
are
Protecting customers from churning
Predicting what customers will
do next
Making decision what kind of services
should be offered to specific customers
|
