工程招标中的围标串标预警模型研究  

Early Warning Model for Together-Conspired or Colluded Bidding Behavior Based on the Similarity of Projects Bidding Documents

作　　者：程铁信[1] 刘文涛 王朝阳 王学海 

机构地区：[1]天津工业大学管理学院,天津300387 [2]天津市水务工程建设交易管理中心,天津300204

出　　处：《天津大学学报（社会科学版）》2018年第1期20-25,共6页Journal of Tianjin University：Social Sciences

摘　　要：工程招投标中的围标串标具有复杂性、隐蔽性及欺骗性特点。模型针对工程投标文件中的结构化数据,采用灰色综合关联分析,计算不同商务投标书之间的相似度;依据投标文件中的非结构化数据,采用文本挖掘的方法,计算投标文件中技术方案等相似度,然后给出综合相似度;并据此判断投标人之间是否存在围标串标行为。工程实例分析表明,该模型既发挥了灰色综合关联分析在小样本数据分析中的优势,又利用了文本挖掘的方法分析投标文件非结构化数据的优势,可以更好地对围标串标行为做出事前判断。Because of complexity,concealment and fraudulence in together-conspired and colluded bidding,the early warn-ing model for together-conspired and colluded bidding based on the big data analysis on projects bidding documents was es -tablished,in which,for structured data in the bidding documents,the Grey comprehensive correlation analysis was applied to calculated the similarity between different commercial tenders,for unstructured data in the bidding documents,text min-ing was applied to calculate the similarity of technical schemes in the bidding documents,then,the comprehensive similari-ty was calculated according to which together-conspired bidding or colluded bidding behavior was judged.The empirical a-nalysis of the bidding case demonstrated that the advantages of Grey comprehensive correlation analysis in the analysis of small sample data and text mining in analysis of unstructured data were fully taken,which could make early warning better for together-conspired and colluded bidding behavior.

关 键 词：招投标 围标串标 灰色综合关联分析 文本挖掘 

分 类 号：C931[经济管理—管理学]