Traffic Engineering Recurrent Spatial Knowledge Base: Design and Implementation

Abstract

Several regional and environmental agencies record spatial data continuously at fixed time intervals such as hourly, daily, monthly, and annually. Such a data collection is termed a recurrent spatial data collection. This paper reports experiences from the implementation of a recurrent spatial knowledge base for traffic analysis. The knowledge base consists of a database, a function library, and a procedure collection. The knowledge is either retrieved directly from the database or computed using functions from the function library. The procedure collection stores various procedures that can be used for statistical and artificial intelligence analysis. The knowledge base organization is transparent to its users. A scientific analysis language is used to provide a spreadsheet-like interface for the knowledge base. Selection of an appropriate data model for the database and its design and implementation is also discussed in detail. The discussion focuses on issues such as knowledge retrieval, access time, and storage requirements. It is argued that the NF