A holistic profile for information filtering systems

Date

1995-05

Journal Title

Journal ISSN

Volume Title

Publisher

Abstract

The quality of information filtering systems is dictated by the quality of the underlying user profile. Most conventional profiles are simply a set of words that describe user interests. While such profiles may be sufficient for casual users, the profile must be much broader in scope before filtering systems prove to be useful in a corporate environment. The creation of existing profiles is unstructured, and may overlook factors that are significant to the decision maker. This paper proposes a holistic profile that addresses this inadequacy by providing a structure to the profile that assimilates not only user interests, but also such considerations as the user's mental model, the user's functional area within the company, concerns related to ongoing projects, client concerns, and contextual information about the organization.

The user's mental model, i.e., his or her internal representation of the how the surrounding world behaves and how to best react to that behavior, is in a state of continuous evolution and can be made richer and more complete by the information gathered by the holistic profile. The holistic profile also enables the decision maker to remain current, informing him or her of opportunities for developing both professional skills and awareness by incorporating features that alert the professional to information that impacts the performance of his or her duties. Ongoing projects can benefit greatly if management or team members are alerted to new developments that might influence the successful completion of the project. Client concerns are a part of the holistic profile, because in order to provide the best client services the user must be aware of external factors that could harm or benefit the client. In addition, because the holistic profile encompasses organizational factors, the information filtering system can serve as an environmental scanning tool for collecting information from the environment to assist in developing strategies that help the organization formulate responses to that environment.

The holistic profile also features domain-specific templates to provide guidance during profile generation, an expansion mechanism to assist in selecting the most appropriate keywords, and a passive learning mechanism to insure that the holistic profile accurately reflects the user's dynamic information needs.

This research has fully developed the notion of a holistic profile and designed, implemented, and tested a prototype "holistic profile filtering system" to demonstrate the increased effectiveness of the information filtering process. Because the resulting profile embraces the areas of professional development, project management, client services, and environmental scanning, it makes information filtering a more versatile tool with increased usefulness in a corporate environment.

Description

Citation