Java Reference
In-Depth Information
functions and algorithms supported by JDM were then introduced to
give the reader a high-level sense of the domain. These mining func-
tions are capable of addressing a wide range of problems. Since Part I
focused on strategy, the specific strategic and tactical elements
behind JDM were elaborated. Lastly, an example was given that
leveraged both the CRISP-DM process and the JDM standard inter-
faces. This provided greater insight into the data mining process and
use of JDM within that process.
Part II focused on the JDM standard itself, describing concepts
from data mining generally, as well as JDM specifically. This was
introduced through a series of examples organized by mining func-
tion. Becoming comfortable with data mining concepts is the first
step toward being able to work with the technology. JDM concepts
help the reader understand how JDM supports automation of, as
well as detailed control over, the mining process. To assist JDM
implementers and users to understand some of the JDM design prin-
ciples, we discussed topics ranging from package design and object
factories to exceptions and discovering data mining engine (DME)
capabilities. Although JDM is focused on Java—targeting the Java
API—we also explored the XML Schema for representing JDM
objects and data mining Web services.
Part III brought data mining into practical focus, explaining how
JDM benefits the application developer and how to use JDM to
build applications. Through a series of full code examples, the use
of the JDM API involving several data mining functions was illus-
trated in the context of specific business problems. An example of a
data mining tool graphical interface was highlighted, illustrating
the use of JDM's capability discovery mechanism to achieve code
portability. To encourage the use of web services, two examples
using JDM web services were introduced—one using PHP, and
another based on JAX-RPC. Because data mining has an impact on
the information technology (IT) departments of both large and
small businesses, areas such as computing hardware, data storage
and access, backup and recovery, scheduling, and workflow were
examined. As with any standard, availability of implementations is
key. Part III finished with an overview of architectures and JDM
implementations from Oracle and KXEN. Some guidelines and
insights into JDM implementation issues were discussed.
Part IV opened with the evolution of data mining standards,
looking at PMML, CWM DM, and SQL/MM DM, and their relation-
ship to JDM. The Java Community Process (JCP)—the means by
which JDM came into being—was discussed, highlighting the
Search WWH ::

Custom Search