The following items are my major publications.

  • An AI system that generates plans starting from any combination of 13 persuasion goal types and using 56 speech act types. It uses integer linear programming and compared well with judgments by human test participants:

Pautler, D. 2007. Computable Social Communication. PhD dissertation, University of Hawaii-Manoa.

(dissertation PDF) (presentation PDF) (downloadable source)

  • An earlier, rule-based implementation that used a different set of goals, and fewer speech act types (40). It was able to guide the user’s topic selection, but was never evaluated with test participants:

Pautler, D. & A. Quilici. 1998. “A computational model of social perlocutions.” Presented at COLING-ACL, Montreal, Quebec. (PDF)

  • A system that helped novice users learn Unix by modeling their likely misunderstandings when doing simple tasks:

Pautler, D. & A. Quilici. 1998. “A Computational Model Of Recognizing and Revising Inappropriate Advice.” Presented at the Cognitive Science Society annual conference. (PDF)

  • A system that generated test scenarios for Army tank units sitting in fight simulators. For example, if the test leader wanted to focus on responding to helicopter attacks, the system would find a spot in the test terrain where such an attack might happen, and guide the unit there through a series of plausible commands and theater updates:

Pautler, D., S. Woods, & A. Quilici. 1997. “Exploiting Domain-Specific Knowledge To Refine Simulation Specifications.” Presented at the International Conference on Software Engineering. (PDF)

Trackback URL