BAYESIAN BELIEF NETWORKS IN DEVELOPING A SCHEMA-TO-SCHEMA EXPERT DECISION SUPPORT SYSTEM

摘要

Extensible Markup Language has become a dominant standard for data representation on the Internet due to the existence of their descriptors and schemas. The use of markup declarations is the only standardized method for creating document type descriptors, but it requires huge data type definitions with some additional properties to explain a true meaning. However, the extensible markup language schema is a way to accomplish the same goal using the language itself. This paper proposes a system to perform schema-to-schema mapping using Bayesian Belief Network as a novel intelligence system scheme for intelligent mapping. The new schema implementation based on Bayesian Network intelligent approach is proposed. The implementation issues of this schema decision support system with schema-to-schema data repository and the training algorithm with decision maker are proposed. An example of Schema-to-Schema Xpert Decision Support System pattern schema-to-schema mapping in representations of relational databases and extensible markup language mapping with the empirical results are proposed to investigate the new intelligence system paradigm utilization. We recommend this new system for more real-world applications, such as, system integration and information retrieval. We are expecting to have stable, reliable, and high quality results.