The problem of formulating knowledge bases containing action schema is a central concern in knowledge engineering for AI Planning. This paper describes LOCM, a system which carries out the automated generation of a planning domain model from example training plans. The novelty of LOCM is that it can induce action schema without being provided with any information about predicates or initial, goal or intermediate state descriptions for the example action sequences. Each plan is assumed to be a sound sequence of actions; each action in a plan is stated as a name and a list of objects that the action refers to. LOCM exploits assumptions about the kinds of domain model it has to generate, rather than handcrafted clues or planner-oriented knowledge. It assumes that actions change the state of objects, and require objects to be in a certain state before they can be executed. In this paper we describe the implemented LOCM algorithm, the assumptions that it is based on, and an evaluation using plans generated through goal directed solutions, through random walk, and through logging human generated plans for the game of Freecell. We analyse the performance of LOCM by its application to the induction of domain models from five domains.
One approach to the problem of formulating domain models for planning is to learn the models from example action sequences. The LOCM system demonstrated the feasibility of learning domain models from example action sequences only, with no observation of states before, during or after the plans. LOCM uses an object-centred representation, in which each object is represented by a single parameterised state machine. This makes it powerful for learning domains which fit within that representation, but there are some well-known domains which do not. This paper introduces LOCM2, a novel algorithm in which the domain representation of LOCM is generalised to allow multiple parameterised state machines to represent a single object. This extends the coverage of domains for which an adequate domain model can be learned. The LOCM2 algorithm is described and evaluated by testing domain learning from example plans from published results of past International Planning Competitions.
The problem of formulating knowledge bases containing action schema is a central concern in knowledge engineering for AI Planning. This paper describes LOCM, a system which carries out the automated induction of action schema from sets of example plans. Each plan is assumed to be a sound sequence of actions; each action in a plan is stated as a name and a list of objects that the action refers to. LOCM exploits the assumption that actions change the state of objects, and require objects to be in a certain state before they can be executed. The novelty of LOCM is that it can induce action schema without being provided with any information about predicates or initial, goal or intermediate state descriptions for the example action sequences. In this paper we describe the implemented LOCM algorithm, and analyse its performance by its application to the induction of domain models for several domains. To evaluate the algorithm, we used random action sequences from existing models of domains, as well as solutions to past IPC problems.
This paper addresses the problem of domain model acquisition from only action traces when the underlying domain model contains static relations. Domain model acquisition is the problem of synthesising a planning domain model from example plan traces and potentially other information, such as intermediate states. The LOCM and LOCMII domain model acquisition systems are highly effective at determining the dynamics of domain models with only plan traces as input (i.e. they do not rely on extra inputs such as predicate definitions, initial, final and intermediate states or invariants). Much of the power of the LOCM family of algorithms comes from the assumption that each action parameter goes through a transition. One place that this assumption is too strong is in the case of static predicates. We present a new domain model acquisition algorithm, LOP, that induces static predicates by using a combination of the generalised output from LOCM2 and a set of optimal plans as input to the learning system. We observe that static predicates can be seen as restrictions on the valid groundings of actions. Without the static predicates restricting possible groundings, the domains induced by LOCMII produce plans that are typically shorter than the true optimal solutions. LOP works by finding a minimal static predicate for each operator that preserves the length of the optimal plan.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.