Thumbnail: click to enlarge

KDWizard Applications

The KDWizard initiative focused on designing, developing, and validating an adaptive framework for knowledge discovery. This framework provides mechanisms for generating knowledge-discovery applications beginning with end-user specifications of decision objectives and sets of data sources. This technology also provides users with a larger and more focused selection of decision alternatives and, consequently, aids them in realizing a more efficient and successful decision making process.

In the initial phase of this initiative, KBSI developed and demonstrated a proof of concept Knowledge Discovery Wizard (KDWizard). The second phase focused on developing the software and demonstrating its benefits on focused military and commercial applications.

Key innovations include the following items.

	Novel approach to designing, generating, and applying hybrids of multiple data mining techniques to support knowledge discovery starting from user goals and raw data sources.
	Innovative ways to combine data-driven machine learning techniques with knowledge based techniques to support knowledge discovery.
	Comprehensive life cycle support for the entire knowledge discovery process.
	Plug and play knowledge discovery toolkit architecture that can be rapidly configured to varying application situations.
	Innovative self-adaptive mechanisms that will allow the knowledge discovery tools to automatically improve their performance over time.