2.4 Story Four: Functional Ontology – Video Structure Meaning

is early computational analysis of the two marathon videos demonstrates the possibilities
of machine-augmented determination of structural attributes and key image frames in a manner
capable of providing adequate power for discrimination and analysis. Presenting the points of dis-
continuity to users is analogous to a topographic map. Various attributes at varying levels of detail
are made evident. e idiosyncrasies of extraction, evaluation, and verbal tagging are left largely to
the user.
e attributes described here are by no means the only attributes of moving image docu-
ments that might be of interest to users, nor are they the only attributes subject to precise descrip-
tion in a manner similar to that used here. Other attributes such as color values and luminance may
not only be of interest but may also provide the means for more fully automated structure analysis.
e audio tracks that accompany some moving image documents also present challenges and
opportunities. In all these cases the numeric and formulaic processes for tracking discontinuities
would remain the same.
Deriving the virtue and power from moving image documents is signicantly enhanced
by the application of digital technology and the modeling it allows. e precise description and
addressing enable book-like access. Linking user navigation through the document to data about
its structure enables user participation in sampling and meaning construction analogous to those
customary in print documents with abstracts, indexes, and forms of marking such as post-it notes
and bending corners. In a way, constructing document contour maps is analogous to an automatic
indexing program for print verbal documents, but because of the complexity of visual documents,
their image based nature, and the potential of the digital environment, contour maps are able to
produce a variety of rich representations which:
are based on the physically present text;
involve the user to some degree;
do not ax concept tags; and
present dierent sorts of image based abstracts.
Contour maps enable navigation in ways familiar from the use of print documents, as well as
ways familiar from navigating physical space. e benets of pre-constructed representations and
idiosyncratic exploration can be combined into a single system.
To return to the title of our piece: the digital environment means theorists and students now have
means to come to grips with the precarious balance between stillness and movement that we encounter
at every step. We can speak closely and carefully about production elements in a way that others will