In spite of their striking diversity, numerous tasks and architectures of intelligent systems such as those permeating multivariable data analysis (e.g., time series, spatio-temporal, and spatial dependencies), decision-making processes along with their models, recommender systems and others exhibit two evident commonalities. They promote human centricity and vigorously engage perceptions (rather than plain numeric entities) in the realization of the systems and their usage. Information granules play a pivotal role in such settings. In the sequel, Granular Computing delivers a cohesive framework supporting a formation of information granules and facilitating their processing. We exploit two essential concepts of Granular Computing. The first one, formed with the aid of a principle of justifiable granularity, deals with the construction of information granules. The second one, based on an idea of an optimal allocation of information granularity, helps endow constructs of intelligent systems with a very much required conceptual and modeling flexibility. The talk covers in detail two representative studies. The first one is concerned with a granular interpretation of temporal data where the role of information granularity is profoundly visible when effectively supporting human centric description of relationships existing in data. In the second study being focused on the Analytic Hierarchy Process (AHP) used in decision-making, we show how an optimal allocation of granularity helps facilitate collaborative activities (e.g., consensus building) in group decision-making.
展开▼