A proposed approach to metadata modeling

Abstract

Metadata is considered a very useful and valuable component in handing data and delivering decisional information, in managing contents on the Internet, in databases and warehouses, in enabling the finding of relevant information, and in effectively providing a multiplicity of useful services. Data, in point of facts, can be metaphorically compared to a dead substance; and all acknowledge the fact that data, by merely existing in systems, represents almost nothing. However, it is the awareness of the exact length and value of this data that builds the valuable expensive information. This thesis presents a detailed overview of metadata and the existing approaches used in metadata modeling, then proposes a new approach rather deviated from the common standards and Web approaches already used; while the previous ones address specific problems, the quest in this thesis goes for a global approach that would suit the largest possible set of applications and scenarios, help in discovering this magic recipe of turning lifeless uninteresting data into valuable active metadata information, and by this define a grand unified metamodel that can be called an Information Model. Thus, having tackled different sides of the available metadata approaches, the objective develops into obtaining - in addition to a comparative study between the existing and the proposed approach - several advantages of the latter one, beginning with a representation of its step-by-step metadata construction methodology, its integration of web and applications metadata, the incorporation of business rules into it, the use of the Einstein Simplicity Theory of dividing and regrouping elements during its conception, and finally ending up with a detailed example to support it, and following the proposed procedure in its implementation.