Artificial intelligence techniques for modeling database user behavior

Abstract

The design and development of the adaptive modeling system is described. This system models how a user accesses a relational database management system in order to improve its performance by discovering use access patterns. In the current system, these patterns are used to improve the user interface and may be used to speed data retrieval, support query optimization and support a more flexible data representation. The system models both syntactic and semantic information about the user's access and employs both procedural and rule-based logic to manipulate the model.