/_   \_
T_
"Navigating between Risk and Return"
 A_
  L_
   E__
  (R)E_
    T_
 _ A_
M_
 

STALEMATE KDD Lab:  
Knowledge Discovery and Data Mining on the World-Wide Web


STALEMATE
  KDD Lab



The STALEMATE KDD Lab studies and provides a testbed for integrated knowledge management-oriented, Web-enabled and pervasively user-interfaced e-business applications featuring:

  • Search engines and data mining on the Internet and intranet
  • Data warehousing and reporting/OLAP
  • Knowledge-based systems and simulation
  • Intelligent browser-enabled user interfacing
  • System integration
  • This site, presently under construction, will subsequently present papers covering the following topics along with illustrative test applications:

    * General Background

    In order to integrate technology-oriented "Knowledge Discovery and Data Mining" with organization-directed "Knowledge Management", a tailored definition of KDD is proposed. The inherent general operational model is subsequently utilized to found a development paradigm for ergonomically enhanced, networked knowledge-management support systems. This development model is considered below, along with technology transfer from in the supply chain upstream scientific research and orientation toward the downstream business and consumer market.

  • Implications of KDD for a Methodology of System Development
  • The development of integrated target knowledge-management applications is considered specifically in the case of strategic knowledge bases for modeling and simulation, model verification through reporting/OLAP for tactical data warehouses and datamarts, and administrative search engines for mining raw data. The envisioned total system-development trajectory involved may be characterized as top-down design with layer-wise bottom-up implementation, simulation-oriented, data-driven, geared toward automatic program generation and open interfaced, lending itself to RAD and prototyping within a corporate enterprise.

  • Sociology of Information Science:
    Towards a Scientific-commercial Transfer Paradigm
  • Upon comparing the scientific research institute with the business enterprise in terms of organization, goals, horizons and culture, the provisions for a synergetic relationship between the two as leading to commercial cutting-edge technology are considered.

  • Marketing Reconnaissance:
    Mass Absorption of Information and Communication Technology
  • Following an examination of the horizontal organization of primary, secondary and tertiary suppliers within the range of vertical markets which they represent, a model is suggested for grading the impact of innovation on the consumer.

  • References
  • * Data Mining and Web Search

    'Web mining' is construed in terms of definition, recognition, controlled continuous search, abstract indexed storage, query and intermediate organization of document and resource information. This initial phase represents the processing of raw (ancillary) materials for knowledge acquisition with exploration taking place on the Internet as well as the intra- and extranet.

  • Search and Location of Resources in Harvesting the Web
  • Indexed Storage of Web Resources in Databases
  • Ontology-based Directories for Databases of Web Resources
  • Directory-oriented Query of Web-resource Databases
  • References
  • * Data Warehousing and Reporting

    Data extraction, cleaning and synchronized storage in integrated, heterogeneous, distributed databases, along with retrieval for reporting and analysis are covered. This intermediate phase represents the construction of datamarts for knowledge testing.

  • Extraction of Compound and Derived Information via Directory-oriented Web-resource Databases
  • Information Storage in Heterogeneous, Distributed Databases based on an Integral Data Model
  • Information Retrieval from a Datamart for Reporting, Analysis and Process Control
  • References
  • * Knowledge-based Systems and Simulation

    A knowledge-acquisition paradigm is deployed entailing a knowledge base of assumptions, rules and assertions to design and implement a model, perform verification through matching with 'real-world' analyses based on a datamart and provide for predictive analyses to support planning. The class of knowledge-based systems involved provides tools for managing knowledge, thereby facilitating discovery, testing, communication, acceptance, application, maintenance and evaluation.

  • Knowledge Base Design as a Formal Methodology for Modeling
  • Implementation of Simulation Programs Representing Knowledge Bases
  • Testing Knowledge Bases as Simulation Programs using Queries over Datamarts
  • References
  • * Web-enabled Intelligent User Interfacing

    Proceeding from ergonomic requirements imposed by knowledge managers on software tools, embedding two levels of intelligence within the user interface is considered.

  • Intelligent Man-machine Communication in Knowledge Management
  • The value of enabling recognition and synthesis of textual, written and oral language along with non-verbal, visual expression within user interfaces for knowledge-management applications is studied.

  • Augmenting User Interfaces with Virtual Reality and Adaptive Agents
  • The embedding in a user interface of an agent or collaborative collective of agents capable of learning within a virtual world is investigated.

  • References
  • * Strategic System Integration

    Effective implementation of knowledge-management applications requires in the course of integration within the organizational context, integration within the relevant technical infrastructure. The interfacing of modular applications across platforms, within intranets and across the Internet, is discussed. An OSI-based application-implementation approach is subsequently considered.

  • Integrated Interfacing of Data-mining, Data-warehousing and Knowledge-based Systems
  • Integrated Implementation of Distributed, Web-enabled Applications to Support Knowledge Management
  • References

  • Kindly submit and review specific comments and questions regarding this page via the guestbook or contact us by email.