A harmony and disharmony in mining of the migrating individuals

摘要

This paper deals with the object detection, recognition and categorization taking into account its ability to convert into some other object. This is possible when the object variance includes an individual ability to change one or more of its properties within some fixed computational framework. The potentially convertible individuals are gathered into the variable domains, in turn. The variance of domains is also considered within a fixed computational model. The construction of a computational model suitable to capture the semantics of variable individuals is proposed as a special `conceptual stand'. This leads to a primitive frame to detect the individual migration from domain to domain. The detection of such a convertible individual gives rise to the task of `mining' the disguised individuals with changed properties. The convertible individuals are modeled by the elements of variable domains. The supporting computational model is based on a separation of individuals into actual, possible and virtual ones. The convertible individual in a model can be recognized by its decomposing into its dimensional counterparts. As was shown, this leads to an adoption of the stage-by-stage cognition model with a pair of evolvents to capture dynamics of the domains - the 2-dimensions model. The first evolvent reflects the generation of the individuals in a domain, the beginning of and canceling out their existence in a domain. The second evolvent reflects the shifts in properties of the individuals. As awaited this unified data model will have the applications to a wide range of models in computer science and Information Technologies.