报业广告智能决策系统的研究与应用

针对报业广告决策的特点,运用数据仓库技术和决策支持系统的思想开发出一套适应面广、灵活性高的智能决策支持系统。基于多层次决策应用的需要提出了虚拟集市的概念和一种改进的三层数据集市结构,并利用星型图数据建模技术将其成功应用到实际系统中;采用基于数据集市和联机分析处理的集成化智能决策支持系统模型,详细论述了系统的预测和客户分析功能模块并提出了基于层次分析法的客户忠诚度定量模型。该系统已成功应用到实际中,对报社广告部的决策提供了有力的支持,取得了良好的效果。

Study and Applications of Newspapering Advertisement Decision Support System

An intelligent decision support system with good adaptability and flexibility for newspapering advertisement is developed by means of data warehouse technology and DSS. With the development of DSS, it is necessary to integrate DW as data source. To the multilevel decisions in newspaper office, a new concept of virtual mart and an improved three-level framework of data mart are presented. Star schema is a universal and feasible method for construct data mart. It is used successfully in practice. The system adopt a compositive IDSS pattern, which is based on DM and OLAP technology. The modules of forecast function and custom analyse function are discussed in detail. A quantitative model of custom persistency based on the analytic hierarchy process is proposed. The application of NA-IDSS archives prospective purposes.