Abstract:
LCW mechanism is one way to handle incomplete information in planning, that was first
investigated by Etzioni et. al. and later advanced by Bacchus and Petrick. This work is
another approach to LCW information in planning based on logic programming. We
describe how LCW information is represented and inferred under logic programming
framework. We also provide a high-level action language AL to capture all properties of
actions that involve LCW information. The focus of defining this language is on semantics
of knowledge updating at planning time and at execution time. Our formalism is shown to
be sound with respect to reasoning about knowledge and actions.
