I'm excited to announce that my paper with R. Michael Young, "Toward Combining Domain Theory and Recipes in Plan Recognition" was accepted for publication at the 2017 Plan, Activity, and Intent Recognition Workshop at this year's AAAI Conference on Artificial Intelligence.
In this paper, we proposed to combine two existing (symbolic) plan recognition variants into a new hybrid variant (but we didn't call it "Hybrid Plan Recognition" since that name represents a combination of probabilistic and symbolic reasoning for plan recognition). The two kinds of plan recognition that we combined were recipe- (library-) based plan recognition and planning- (domain theory-) based plan recognition. Recipe-based approaches are more expressive but require more manual input, whereas planning-based approaches are less expressive but more general (i.e. require less input and do more). We therefore proposed a technique "to retain the flexibility, generality, and scalability of the domain theory-based approach and the representation of typical non-optimal action sequences afforded by the library-based approach."
We present a technique to further narrow the gap between recipe-based and domain theory-based plan recognition through decompositional planning, a planning model that combines hierarchical reasoning as used in hierarchical task networks, and least-commitment refinement reasoning as used in partial-order causal link planning. We represent recipes through decompositional planning operators and use them to compile observed agent actions into an incomplete decompositional plan that represents them; this plan can then be input to a decompositional planner to identify the recognized plan-space plan. Our model thus synthesizes the heretofore disparate recipe-based and domain theory-based plan recognition variants into a unified knowledge representation and reasoning model.
Rogelio E. Cardona-Rivera and R. Michael Young; Toward Combining Domain Theory and Recipes in Plan Recognition. In Proceedings of the Plan, Activity, and Intent Recognition Workshop at the 31st AAAI Conference on Artificial Intelligence (PAIR 2017), pages 796-803, San Francisco, CA, USA, 2017.