一种基于CBR的网络水产品价格预测方法  

A CBR-BASED APPROACH FOR FORECASTING WEB PRICES OF AQUATIC PRODUCTS

作　　者：袁红春[1] 陈明[1] 黄冬梅[1] 

机构地区：[1]上海水产大学信息学院,上海200090

出　　处：《计算机应用与软件》2008年第12期39-42,共4页Computer Applications and Software

基　　金：国家"十一五"科技支撑计划项目(2006BAD10A11)

摘　　要：水产品价格的科学预测对水产业健康可持续发展具有重要作用。提出了一种基于案例推理CBR(Case-Based Reason-ing)的水产品价格预测方法,其包括网络数据自动获取、基于概念树的面向属性归纳、案例的生成与表示、案例匹配及相似性计算、案例评价与修正等关键过程。在关键过程研究基础上实现了预测系统,并对网络水产品价格数据进行预测实验。结果表明,该系统能自动采集权威网站水产品价格数据,并能对水产品价格进行有效的分析与预测。It plays an important role to scientifically forecast the prices of aquatic products for the healthy and sustainable development of aquaculture. In this paper it presents a method for forecasting aquatic products＇ prices based on Case-Based Reasoning （CBR） and its key processes, such as automatic web data acquisition, attribute-oriented induction based on concept tree, generation and representation of eases, cases matching and similarity calculation, case evaluation and revision, etc. A prediction system based on the key processes study has been implemented, and forecasting experiments for price data of web aquatic products have been conducted. Results show that the system can automatically collect web data from authoritative websites that issue aquatic products＇ prices and can effectively analyze and forecast the aquatic products＇ prices.

关 键 词：水产品价格 预测 面向属性归纳 基于案例推理 

分 类 号：TP393.08[自动化与计算机技术—计算机应用技术] F726[自动化与计算机技术—计算机科学与技术]