Knowledge Nebula Crystallizer

"Knowledge Nebula Crystallizer (KNC)" has been originally suggested by Hori et al.¹⁾ as a prototype knowledge management system which has a repository called "knowledge nebula". The knowledge nebula is an unstructured collection of small information pieces. Essential operations of the KNC are crystallization and liquidization (see Knowledge Liquidization & Crystallization).

During the crystallization, information pieces from the nebula are selected and structured according to a particular context, resulting in a new information artifact. During the liquidization, an information artifact is chunked into elements which are added to the knowledge nebula.

Knowledge Nebula

An unstructured collection of small granularity

Knowledge Nebula Crystallizer

A system which conducts Knowledge Liquidization & Crystallization, and a repository for Knowledge Nebula

The Process of Knowledge Creation

The KNC is regarded as a co-authoring tool. The KNC conducts what human conducts in his/her mind. Briefly, the process of knowledge creation proceed in the way shown in the picture below:

A designer refers to both outside information resources and that in the designer's mind. The designer excerpts portions of information from both of them (the process of knowledge liquidization) and fuses into an information piece (the process of knowledge crystallization).

The KNC aims to support this process by providing the designer with a place for reflective thinking with functions that computers are good at, such as searching and retrieving. That is, the KNC supports the designer by externalizing the designer's mental space and stimulating his/her mental space.

The process of co-authoring is as follows:

• A designer presents an information artifact, such as keywords, a piece of music / video, a paper that he/she wrote in the past.
• Then the KNC reads the information artifact and provides some possible forms of knowledge crystallization as stimulants from the knowledge nebula.
• Through browsing / analysing the presented candidates, the designer crystallizes his/her knowledge for improving the information artefact.
• While the designers are interacting with the information artefact, the KNC reads both the information artefact and the interaction and provides some possible forms of knowledge crystallization as stimulants again.
• When the designer completes the information artefact, he/she puts it into the KNC and the KNC liquidizes it into the knowledge nebula.

The picture below depicts this cyclic process:

Through this cyclic process, both human mind and the KNC evolves. "Dynamic Concept Base (DCB)" is implemented inside of the KNC that defines similarities among information pieces in the knowledge nebula. The similarities are modified dynamically through interactions between a designer and the KNC in order for the KNC to provide proper forms of knowledge crystallization to the design context. The picture below shows the image of the DCB.

The DCB not only modifies the similarities dynamically, but preserves plural definitions of similarities in order for users to re-use in the future similar context.

I have been challenging to put theoretical frameworks for knowledge creation into practice, i.e. apply the KLC　to practices. In order to apply KLC & KNC to practices, the following questions has to be answered:

• How should Knowledge Liquidization be conducted?
• What should the knowledge nebula be?
• How should Knowledge Crystallization be conducted?
• How should the stimulants be represented to a designer?
• What interactions are required?

Currently I am developing an KNC for time-based and non-verbal information authoring, such as musical and video compositions. Details are to follow.