Alessandro Mosca

Research in the Digital Humanities requires computer systems that can document and analyze interpretations of creative works, such as literary texts. These systems are designed to provide access to existing interpretations and identify their similarities and differences. To this end, we propose an approach to formally representing interpretations. In particular, we develop an ontology of observations aimed at capturing, through what we call observational vocabularies, what interpreters claim about the texts that they interpret. We distinguish between different types of observations, in particular, basic observations, which represent specific and domain-dependent claims (e.g., regarding the analysis of literary characters’ traits), and observations such as assertion, denial, support, and defeat to document more nuanced claims. We also introduce formal mechanisms that can be used to analyze particular (sets of) observations, texts, and so on. Throughout the article, we illustrate the discussion with examples from literary studies.

The main aim of this paper is to present a way to improve the e.ciency of an intelligent alarm correlation module by means of an abductive reasoner. The alarm correlation module is integrated in a more complex system that performs monitoring and control tasks over a tract of a highway. On the basis of a specific theory on the domain, explanations of anomalous traffic patterns can be provided taking into account those situations not directly detected by data acquisition technology. The integration of the comprehensive view obtained by means of a correlation phase with additional domain knowledge allows to abduce from the observed anomalous traffic patterns other significant situations. The experience of traffic operators allows to build a theory of abducible explanations that exploits the knowledge of the use context of the system. The correlation module of SAMOT has been modelled within a propositional modal language, named ST-Logic, which will be adopted for the abductive reasoner as well. Finally, some remarks about stratified interpretative cycles made possible within such a formal framework will be addressed.