Year: 2009
Type: conference-paper
Abstract: This thesis work concerns the area of automated acquisition \nof planning domain models from one or more examples of \nplans within the domain under study. It assumes that an adequate \ndomain model for a domain can be composed of objects \narranged in collections called object sorts. Recently, two \nsystems have had success in using this underlying assumption: \nthe Opmaker2 system (McCluskey et al. 2009), and \nthe LOCM system (Cresswell, McCluskey, and West 2009). \nThe former requires only one solution plan as input, as long \nas it contains at least one instance of each operator schema \nto be synthesized. It does require a partial domain model as \nwell as the example plan, and the initial and goal states of \nthe plan. In contrast LOCM requires no background information, \nbut requires many instances of plans before it can \nsynthesize domain models. Our aim is to build on these systems, \nand establish an experimental and theoretical basis for \nusing object - centred assumptions to underlie the automated \nacquisition of planning domain models.
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W1218646