Plan Applier Mechanism

This lecture from a course on Artificial Intelligence focuses on the Plan Applier Mechanism (PAM), a program developed by Roger Schank and his students in the late 1970s for story understanding. PAM uses knowledge about goals, plans, and actions to make inferences and build an explicit representation of a story. The lecture explains the three phases of PAM: predictive, bottom-up, and incorporation. In the predictive phase, PAM tries to match input conceptualizations with expectations. If no match is found, it enters the bottom-up phase to generate explanations, and finally incorporates the results into the story representation. The lecture provides a detailed trace of PAM processing a story about John wanting Bill’s bicycle, showing how PAM explains actions by linking them to plans and goals, and how it handles goal conflicts. It also demonstrates PAM’s question-answering capabilities and its ability to generate different character perspectives. The lecture concludes by summarizing the three phases and indicating that the algorithm will be discussed in more detail in the next class.