Purpose Computation-Oriented Modeling and Transformation on DIKW Architecture

In recent years, a foreseeable AI landscape with explainable and interactive human interactions is becoming feasible based on DIKW premises. The DIKW modals are increasingly acknowledged as an important approach to address the problems related to semantic understanding beyond various question and answering systems. However, there continues to be no unified understanding over the meaning of the DIKW concepts. Data, information, knowledge, and wisdom, as a whole concept of DIKW, are also missing cohesive understanding of the relationships among them. This chapter splits the incomplete, incompletely correct, and imprecise resources into data resources DDIK, information resources IDIK, and knowledge resources KDIK according to the DIKW architecture. It narrows the complexity of resource processing and further clarifies the definition of DDIK, IDIK and KDIK. We regard information as the state after the special purpose is given to the data resource, and the specific purpose of human is the key point of data and information transformation. Therefore, we propose the new purpose resources PDIKs from the typed resources, which explain the specific purpose of human beings. Based on the existing internal relationships of these four kinds of resources, the chapter constructs the corresponding typed resource systems (DSystem, ISystem, KSystem, PSystem, and DIKPSystem) and models the systems to obtain the corresponding typed resource models (DModel, IModel, KModel, PModel, and DIKPModel). We use some examples and diagrams to clearly express and illustrate the DIKPSystem and DIKPModel and the conversion processes of DDIK, IDIK, and PDIK, which have opened up the channel of data and information transformation. Associating DDIK, IDIK, and PDIK, and combining the conversion processes and the tree systems of typed resources, can explain the abundant semantic content and reduce the storage space of resources.