Analyzing conflicts with concept-based learning

Abstract

A machine learning technique for handling scenarios of interaction between conflicting agents is suggested. Scenarios are represented by directed graphs with labeled vertices (for mental actions) and arcs (for temporal and causal relationships between these actions and their parameters). The relation between mental actions and their descriptions gives rise to a concept lattice. Classification of an undetermined scenario is realized by comparing partial matchings of its graph with graphs of positive and negative examples. Developed scenario representation and comparative analysis techniques are applied to the classification of textual customer complaints. © Springer-Verlag Berlin Heidelberg 2005.