Hybrid Subjective Decision Support System based on Computational Semiotics and Computational Intelligence Techniques

Subjective decision problems involve personal feelings and opinions, adding substantial complexity to evaluate different candidate options. In order to deal with this kind of problem, in which individual experience is considered and impacts directly in decision making process, many computational methods have been applied. However, the traditional approaches are often not flexible to consider uncertainties, imprecise situation contexts and idiosyncrasies. In this sense, we present a Hybrid Subjective Decision Support System based on Computational Semiotics and Computational Intelligence techniques. Our approach relies on Case-based Reasoning as the problem solving main methodology and Self-organizing Maps, which acts as pattern recognition tool, in order to organize more appropriately retrieval of similar past cases. Furthermore, a semiotic model handles a prior knowledge (i.e. knowledge acquired from a specialist) and domain specific restrictions to guide the search process towards an appropriate problem solution. In the paper we illustrate how the proposed approach can deal graciously with subjective concepts providing a more intuitive and evident decision making.