Browse Prior Art Database

Intelligent Miner

IP.com Disclosure Number: IPCOM000118456D
Original Publication Date: 1997-Feb-01
Included in the Prior Art Database: 2005-Apr-01
Document File: 6 page(s) / 146K

Publishing Venue

IBM

Related People

Agrawal, R: AUTHOR [+4]

Abstract

Disclosed is a method of an Intelligent Miner architecture and a brief description of the individual components.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 42% of the total text.

Intelligent Miner

      Disclosed is a method of an Intelligent Miner architecture and
a brief description of the individual components.

      A data mining operation is typically made up of several
distinct steps combining the execution of data-mining kernels with
data processing functions.  The user of the data-mining application
wants to be able to select and control the execution of these
functions in a flexible way.

The Intelligent Miner provides the means for this selection and
control, and it features:
  o  A set of data-mining kernels providing the most frequently
      required data mining technologies.
  o  A processing library providing functions like bulk load of
      data and data transformation.
  o  A flexible way for read and write access to flat files and
      relational database tables.  This Application Processor
      Interface (API) is only available to toolkit internal
      functions, not to applications.
  o  A set of API functions to control the execution of the data
      mining kernels.
  o  A set of API functions to manage the results of data-mining
      runs.
  o  A client/server structure to provide communications between
      data mining and data processing functions on the server, and
      administrative and visualization functions on the client.
  o  An Administration Graphical User Interface (Administration
      GUI) to control the execution of the data-mining kernel and
      processing functions and to manage the results of the
      data-mining runs.

The main components shown in the Figure are:
  o  Administration Graphical User Interface (Administration GUI)
     The Administration GUI provides the means for an end user to
      specify input, output, and control parameters for the mining
      kernels and data processing functions, as well as the
      management of results.
  o  Processing Library
     The Processing Library allows the user to collect and prepare
      the input data from various databases.
  o  Environment Layer API
     The Environment Layer API provides a set of data types and
      classes to control the behavior of the data-mining kernels
      and data processing functions.  These classes are made
      available as C++ classes that allow you to build applications
      combining different functions.
  o  Client/Server Component
     This component addresses the communication between data
      types, client classes, and server classes.  It is implemented
      using Remote Procedure Calls (RPC).

    Data Mining kernels

    Kernels are available for:
       - Discovery of Associations
       - Discovery of Sequential Patterns
       - Discovery of Similar Time Sequences
       - Prediction of a Classification
         - Tree Induction
         - Neural Induction
    ...