Â¥Ö÷

Ò»Ö±¾õµÃÕâ¶Î»°²»´í (TopQuadrant Technology Briefing, Semantic Technology, Version 1.1, September 2003) How Knowledge Models are different from other Software Models A model describes how concepts and phenomena are similar, and how they differ, what is called the commonality and variability of concepts in a chosen area of interest, sometimes also referred to as a domain of discourse. The most commonly used models in software engineering are object and data models: The Object Model in an object-oriented program is a networked data set that describes the system itself. In an object model, classes high in the hierarchy express properties that are shared by many system elements; classes low in the hierarchy describe properties that are specific to small sets of elements. Therefore, it is a model that reflects and describes properties and functions of a specific system. The Data Model describes the world outside of the system. Many applications can share the same database, but in reality the schema of the data is typically fine-tuned to the needs of specific application. In a data model, each table in the schema dictates what this collection of records has in common; another schema denotes this for other records. Differences are represented both by individual records, as well as record types. The relationships are held in special index tables and are not explicitly defined. Semantic models are intended as a way for different agents (applications and/or people) to interoperate and to share meaning. Unlike object models they describe the world that is outside of any of the application that uses the model. Furthermore, the variations and commonalities semantic models represent are not of a single entity or stakeholder. By definition semantic models support multiple viewpoints. This makes them especially suitable for solving interoperability problems.