It's a Built-In or Bolt-On Goldmine for Knowledge Discovery Applications" - Announcing K.wiz, Data Mining Technology for ISVs Building Analytical Applications
ACTION ITEMS

Compression Sciences (a subsidiary of Gentia Software) has announced the launch of K.wiz, modern and powerful data mining software.

K.wiz is entirely written in Java and is already compatible with today's release of the Java 1.2 platform. It leverages investments in Web, data warehousing and OLAP systems by revealing the underlying patterns in data and presenting them in a clear, graphical format. K.wiz systems are automated, freeing up human resources and finding answers to questions which are too complex for traditional tools to address.

According to Guy Creese, senior analyst at Aberdeen Group in Boston, MA, "Independent Software Vendors (ISVs) and Value Added Resellers (VARs) need an easy to use data mining tool kit that automates many labor-intensive data mining processes, scales as databases grow and user populations increase, and deploys to a wide variety of users and platforms. Written in platform-neutral Java and including an extensive set of Application Program Interfaces (APIs), K.wiz has clearly been designed to address this need."

K.wiz is a data mining solution that allows data for the knowledge discovery process to be held in as little as 1% of its original volume, optimizing virtual memory usage. Unique "look-up" methods deployed in the compression technology allow results to be accessed many hundreds of times faster than in the original source.

Advanced graphical techniques are used, such as 5D scatter diagrams, allowing relationships in data to be easily identified.

"Visualization is the key to understanding complex relationships in the data," continued Creese. "K.wiz delivers not only intuitive 3D graphics but an animated data mining plan that brings the process alive and allows immediate feedback on the progress of the data mining process."

The software can be used by ISVs to build analytical applications and is expected to be incorporated into many data warehousing and business intelligence solutions.

"Put simply, K.wiz is a built-in or bolt-on goldmine for anyone looking to embed knowledge discovery in their analytical applications," said Ian Rawlings, managing director at Compression Sciences. "By combining object intelligence with total ease of use, companies will be able to identify new market opportunities and exploit competitive advantage."

"K.wiz is developed using Java-based, scalable technology which can be deployed easily across the enterprise," continued Rawlings. "For the first time, end-users will be able to leverage all the benefits of data mining through their usual software interface."

Software developers searching for a data mining engine will be able to embed K.wiz into their own applications. This will enable end users to run the software to identify market trends and predict customer reactions that would normally be hidden.

Unlike traditional data mining solutions, K.wiz has been designed for the application developer by allowing access to every function of the product through programmable APIs. Entirely Java-based, it handles both multi-platform and distributed client/server environments.

"The fact that the K.wiz architecture has been designed to allow the embedding of third party algorithms enables it to be extended to address almost any application requirement," said Gavin Meggs of British Telecom. "Being able to identify and schedule the availability of servers to K.wiz, and then specify components of the data mining process to run on targeted machines, allows organizations to exploit their investment in network infrastructure, and ensure the data mining process runs at optimal efficiency."

The key benefits of K.wiz for ISV developers include:

• Ease of deployment
  K.wiz runs across all common business environments for ease of use. By including advanced communications, models and results can be easily distributed across a network or intranet.
• Automated processes
  Common processes, such as data preparation, can be grouped together for re-use into a user defined custom component that can be leveraged by all K.wiz users, enabling higher productivity for ISVs.
• APIs
  The product has been developed with ISVs in mind. It takes a modular approach, using APIs to allow access to all relevant parameters and K.wiz functionality.
• Scalability
  K.wiz is designed to avoid traditional data mining problems, where applications do not scale to department or enterprise level. By incorporating patented Array Indexing & Minimum Entropy Database (AIMED) technology in its cache and K.wiz/ADB (Accelerated Database) in its Knowledge Mart, data volumes are small and access times are kept to a minimum Additional Information Compression Sciences (a subsidiary of Gentia Software) is a specialist in knowledge discovery solutions. Founded in 1996 and with offices in Boston, MA and the UK, its aim is to enhance business efficiency by automating the on-line delivery of information discovered from massive databases.

Compression Sciences' principal product, K.wiz, is a fully integrated solution designed to provide business end users with powerful analysis capabilities. K.wiz can be delivered through host applications, offering the end user data mining functions with no need to learn specialist applications. Unlike traditional solutions, K.wiz has been designed for Web deployment, enabling easy distribution to large numbers of users.

The key spokesman for the company is Ian Rawlings, managing director. He has over 15 years experience in the field of management techniques, including a commercial background in the development of data warehousing and OLAP systems.