Working with long texts over a long period of time places a heavy cognitive load on an author of such work. One needs to keep in mind many facets of each character. What is her hair colour? What colour are his eyes? Where is the character at which point of the story? Wouldn’t it be great to ask this system to highlight all places in the text where–let’s say–Joe and Susan have scenes together or where Susan has a scene with someone else. It then would be much easier to rewrite scenes (e.g., replacing one character with another one or putting the characters to another location). Moving in the manuscript by means of semantic querying and browsing, and not by syntactical means.
Musing about Semantic Web applications I could easily envision some tool that helps keeping track of such details and helps keeping the respective universe consistent. This idea is almost certainly not a new idea. But maybe the Nepomuk system will provide such features using the text analysis components together with Leo’s Rebith machine?
Well, we’ll see, won’t we?