Abstract: |
In this talk I will give an outline of my doctoral research, which is
now in its early stages. I will motivate and define the problem of
modeling narratives as semantic networks, with particular emphasis on
capturing notions of creativity, reducing dependence on knowledge and
mimicking the human narrative instinct. Touching on prior work, I will
propose processes by which an "expectation model" for narratives will
be statistically learned from common patterns discovered among training
narratives, and discuss the utility of such a model. Finally, I will
outline the progress I have made, including a specific model I am using
to model a subset of Aesop's Fables, as well as the toolkit I have
developed for implementing it.
|