Decision making with natural language based preferences and psychology-inspired heuristics.

Decision making is required by many tasks, such as shopping, nowadays assisted by software systems, and providing support to the decision making process is a feature that would significantly improve such systems. Many decision support systems and related approaches have been proposed to that purpose, but they often involve tedious elicitation processes or previously collected data. In this paper, we propose an automated decision making technique, which chooses an option from the set of those available based on preferences and priorities expressed in a high-level preference language, exploiting natural-language terms, such as expressive speech acts. Moreover, in order to make a decision, our technique goes beyond the provided preferences with psychology-inspired heuristics, which concern how humans make decisions, as provided preferences are typically not enough to resolve trade-offs among available options. Two studies were performed to evaluate our approach, and results indicate that our technique is effective both by comparing its recommendations with those made by a human expert, and by considering evaluation scores provided by users that experienced our technique. [ABSTRACT FROM AUTHOR]