Maurizio CATTANI1 / Glauco MANTEGARI2 / Alessandro MOSCA2 / Matteo PALMONARI2
(1University of Bologna, Italy / 2University of Milano Bicocca, Italy)
The main objective of the paper is to present the results of a preliminary work concerning the use of Knowledge Representation and Reasoning (KR&R) formal techniques for the treatment of archaeological stratigraphy. KR&R is that part of Artificial Intelligence (AI) that is concerned with the study of thinking as a computational process. To provide the most appropriate technical tools for tackling specific classes of problems, KR&R exploits both the logic formalisms for the symbolic representation of knowledge and the efficacy of a number of different computational models for the automatic processing of the inference patterns. The symbolic representation of the stratigraphic units and the inferences that are performed on the sequence constitute the two main aspects in the study of archaeological stratigraphy; thus, KR&R may represent an innovative approach in this field.
Moreover, non-monotonic logical formalisms, contrary to monotonic classical logic, supports the representation of defeasible inference, i.e. that kind of inference in which reasoners draw conclusions on the basis of uncertain and/or partial information, reserving the right to retract them in the light of further information. As regards to this point, the present contribution aims at outlining the introduction of a non-monotonic logic for spatial and temporal reasoning on stratigraphy, which allows the co-existence of different chronological interpretations of the sequence.
The resulting system has been implemented within the Answer Set Programming (ASP) paradigm and empirically tested on the Bronze Age excavation in Mursia (Pantelleria, Italy). The preliminary results of the project will be discussed focusing both on the conceptual and on the technological aspects.
Keywords: Stratigraphic models, Knowledge Representation, spatial-temporal reasoning, Answer Set Programming