Browse Prior Art Database

Knowledge filter for Concept Capture tools

IP.com Disclosure Number: IPCOM000015815D
Original Publication Date: 2002-Aug-16
Included in the Prior Art Database: 2003-Jun-21
Document File: 2 page(s) / 46K

Publishing Venue

IBM

Abstract

A body of knowledge can be represented, at least in synoptic form, by a diagram comprising such visual elements as nodes, links and branches. This diagrammatic representation might be called, for example, a semantic map, a mind map, or a brain dump. Disclosed is an interactive system for exploring and enhancing a knowledge base. For this system, the term concept map is used to describe an arrangement of concept nodes interconnected by semantic links, such as "is a kind of". The system comprises a unique combination of functions: A GUI that provides a concept map view of the knowledge base Agents that work with the knowledge base, both autonomously and under user control Collaboration between users, exploiting the GUI and the knowledge base, while permitting users to control access to their individual information.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 52% of the total text.

Page 1 of 2

Knowledge filter for Concept Capture tools

A body of knowledge can be represented, at least in synoptic form, by a diagram comprising such visual elements as nodes, links and branches. This diagrammatic representation might be called, for example, a semantic map, a mind map, or a brain dump.

    Disclosed is an interactive system for exploring and enhancing a knowledge base. For this system, the term concept map is used to describe an arrangement of concept nodes interconnected by semantic links, such as "is a kind of". The system comprises a unique combination of functions:

A GUI that provides a concept map view of the knowledge base Agents that work with the knowledge base, both autonomously and under user control
Collaboration between users, exploiting the GUI and the knowledge base, while permitting users to control access to their individual information.

    Although each of these functions have been implemented elsewhere, the combination provides a uniquely powerful system that enables multiple users to manipulate knowledge simultaneously while maintaining a holistic model and representation of that knowledge.

    The tool presents the arrangement of nodes and links in a GUI that interactively offers a view of the underlying knowledge base. It differs from analogous tools in that fresh concepts entered into the GUI are filtered by agents that interrogate the knowledge base to provide visual feedback about the relationship between the fresh concept and the knowledge base. Users can also instruct agents to enhance the links from and to a given concept, for example by creating a sub-map. In doing so, the agent can interact with one or more of the knowledge bases known to the system, or can interact with external sources on the world wide web.

  There are two approaches to implementing the knowledge base: Using standard relational database technology Using a tuple-spaces approach, as described in pending US patent application 09/834,782, which is the preferred implementation.

    The knowledge base can be distributed over multiple computers, either as a single distributed database, or as a number of discrete databases communicating through agents.

    The tool also provides additional facilities that exploit both the collaboration between users and the filtering capabilities of the agents. For example:

Collaborators can work individually for periods of time while continuing to be informed on a subscript...