A New Form of Custom Publishing:

MediaMatrix is a software system for publishing multimedia information from a fully integrated database. But text as a database element leads to the concept of topical kernels of information, which may be the appropriate way to think of unitized information whether it is expressed via text, video or even experiment simulations and reconstructions. Kernels should consistently make reference to any "higher-level" and "lower-level" kernels of relevance. By higher-level, we mean those which summarize or reflect upon relations which exist among several more detailed topical kernels. Lower-level kernels are those which describe a given singular topic in detail.

When database units are initially constructed by focusing on the lowest-level of detail, this gives rise to all sorts of opportunities for their selective incorporation into highly personalized collections of content. When reflective kernels --i.e., those which reflect on relations among selected detailed kernels -- are missing in the database, personalized integrative essays which reflect on why these certain kernels have been selected and how each relates to other relevant selected topics may be written by the author making the selections. For example, an author reviewing a topical database of kernels describing various medical symptoms might choose detailed essays on body fevers, sore and reddened throats, and inflamed sinuses, only to find that an essay is missing on a given virus which causes such a collection of symptoms. They might have subsequently been used to describe the diagnosis of a cold, a given strain of flu virus, or a specific allergy, depending on the author and his/her purpose for selecting those elements. In each case, the selecting author will want an integrative essay on colds or alleries, or the strain of flu. When they are already available within the database, they serve as the "higher-order" integrative kernel to the selected "lower-level" kernels. When an integrative essay is found missing, the author will probably want to write his/her own interpretation of how each detail fits with the other to give a higher-order view of their inter-related relevance.

When such essays are submitted and approved for inclusion into the database corpus, the database not only grows in depth and breadth, it also allows for immediate and personalized publication of complex collections of information. This may, in fact, play a significant part in defining the future of educational textbook alternatives -- especially in the physical and social sciences. The model just described is actually a variation built upon much the same model as already exists when authors of scientific articles reference and integrate previous works of relevance, either to why a given study was executed or to the interpretation of its results.

Even textbooks are little more than compilations of integrative essays on the relevance of a multitude of referenced studies and theorectical writings. When the core of such information about principles and/or events has already been condensed, and various intergrative comments have been written, custom publishing is merely a matter of selecting and/or adding by personal authorship all of those topics thought relevant to one's educational or communicative purpose. This is truly custom publishing. And electronic delivery via the internet allows both for "just-in-time" and highly personalized selections.

If topics appropriately and consistently mention who, what, where, when, and why regarding the informational content, such kernels might be reorganized as a timeline, a geographically related collection, or even a biographically organized publication. After all, there are only about five alternative ways of organizing informational access -- alphabetical, topical, coding systems (like the Dewey Decimal or Library of Congress systems), by author, or by date or other relevant publisher information.

Likewise, alternative mapping allows for the simultaneous and metaphorical publication of an entire "collection" of alternative multimedia books -- although "books" is hardly an apt description, given that each unit could just as easily be constructed as a united set of inter-related "video" scenes (the equivalent of a textual kernel). Because such flexibility also opens the possibility of allowing authors the opportunity to personally map a given database and to publish that personal view of inter-connectivity between topical kernels, what may eventually be of the most value in a topical database is the personalized mapping various experts make of it!

Next Topic...Kernelized Information