I'm looking for innovative ways to create text summaries and I was wondering whether feature structures can be used for such a task.
Is it possible to create subsume and unification methods that allow one to model text summary tasks using feature structures?