Paper: Individuality And Alignment In Generated Dialogues

ACL ID W06-1405
Title Individuality And Alignment In Generated Dialogues
Venue International Conference on Natural Language Generation
Session Main Conference
Year 2006

It would be useful to enable dialogue agents to project, through linguistic means, their individuality or personality. Equally, each member of a pair of agents ought to adjust its language (to a greater or lesser extent) to match that of its interlocu- tor. We describe CRAG, which generates dialogues between pairs of agents, who are linguistically distinguishable, but able to align. CRAG-2 makes use of OPENCCG and an over-generation and ranking ap- proach, guided by a set of language mod- els covering both personality and align- ment. We illustrate with examples of out- put, and briefly note results from user stud- ies with the earlier CRAG-1, indicating how CRAG-2 will be further evaluated. Related work is discussed, along with cur- rent limitations and future directions.