Briefs

New Version Offers the Latest in Data Mining
Technology with Real-time Deployment Analytics

SAN FRANCISCO, June 27 /PRNewswire/ -- KXEN(TM), a global analytic
software company that provides powerful, predictive and descriptive
modeling technology for both data mining professionals and business
decision-makers, today announced Version 2.0 of its analytics software
suite. The new version includes enhancements in data structure and
architecture improvements, resulting in increased speed and accuracy
to deliver real-time analytics.

Version 2.0 offers significant gains in speed and accuracy compared to
Version 1.0. The new version was benchmarked against Version 1.0 on 33
datasets, and the benchmark resulted in a one to five percent increase
in the KXEN information indicator (KI), and between a three and 10
percent increase in speed. KI measures the amount of information or
predictive power, which can be extracted from a given dataset and it
ranges from zero to 100 percent.

"The enhancements to our suite take us to the next level -- providing
our customers with real-time analytics," said J.D. de Haseth,
president of KXEN. "The addition of the automatic categorization
functionality takes the guesswork and manual labor out of banding
numeric variables, while providing increased accuracy and ease of
interpretation of results. This is significant for business users who
may not have the statistical expertise and time to create these
categories or bands on their own. Due to the market demand for
powerful analytical tools, the modifications to the internal
architecture enable us to meet our partners' requirements to more
quickly deliver seamless products to the market."

KXEN's full suite of patented analytical software is easy to integrate
into applications and specific industry solutions, including customer
relationship management, sales automation, supply chain management,
network management, financial services, manufacturing, and
biotechnology. The KXEN suite automates the data preparation, coding,
and modeling process and enables users to build predictive and
descriptive models in near real-time for better business
decisions. The models can also be deployed in real-time using KXEN's
runtime server. The suite currently includes three core products:

* K2R (KXEN Robust Regression) -- is a high-speed data mining solution
for prediction and classification. KXEN's patented technology uses
specific regression algorithms that build predictive models quickly and
easily

* K2C (KXEN Consistent Coder) -- automates tedious and time-consuming
data preparation and pre-processing

* KCG (KXEN C Code Generator) -- generates a model built by the KXEN
suite into C code

The data preparation architecture was completely redesigned to go from
a mono-target system to a multiple target system. This will allow
users to solve business problems that cannot be formulated by binary
or continuous models. For example, KXEN components in the future will
be able to optimize multiple offers in a campaign, or multiple
products with different price points.

For additional information, pricing and availability, see www.kxen.com