A2A definitions map actions onto complex animations plus a set
of parameters. Parameters depend on the schematic structure of an
action as encoded by the corresponding frame (consulting offline
the FrameNet lexical resource [29] or some process ontology
[30]): for example the action ‘‘walk’’ can expressed by specifying
a path along which the action occurs, the source and destination
locations of the action, the area in which the walking action takes
place. The work described in [31] exemplifies how characters’ actions
can be mapped onto ontology concepts via the lexical mediation
provided by FrameNet.