Analysis Of Collusive Behavior Of Bidders In Construction Projects Based On Complex Network

Bidding system plays a very important role in China's construction industry.Since the system was carried out,bidders ring and colluding bid behaviors has been repeated many times,which has seriously affected the fair and orderly competitive environment of the construction market.At present,China's Internet technology is developing at a high speed,traditional bidding is shifting to electronic,e-bidding provides big data support.Therefore,based on the historical data of bidding behavior,it is of great significance to study the conspiracy behavior in order to reveal the invisible behavior in the bidding process,reduce and supervise the conspiracy behavior,and promote the healthy and orderly development of market transactions.This paper takes the conspiracy behavior of construction project bidding as the research object,and carries out research on the analysis of collusive behavior.On the basis of reviewing relevant laws and regulations and actual cases,the basic concepts of bidding are expounded,the manifestations and causes and influencing factors of conspiracy are clarified.Based on behavior analysis,the complex network of bidders is constructed by using complex network technology,and the community bidding behavior is identified by community division algorithm.Taking the data of bidding candidates for some public construction projects in S Province as an example,this paper makes an empirical analysis on the discovery of bidder's conspiracy associations.Finally,the corresponding supervision mechanism is put forward to control the bidders' collusion in bidding market of construction projects.The main conclusions of this paper are as follows:(1)The analysis model of bidder's complex network and community structure is constructed.The model is based on the data of successful bidders in bidding projects,the characteristics of the network are described by index of degree,weight and centrality.BGLL algorithm and k-cliques algorithm are used to classify communities,and modularity and strong/weak index are set to evaluate the quality of network partition.(2)Takeing the partial winning bidding data of S Province as an example,a real market bidder network is established.The results show that the average degree and weighted degree of the network are 7.507 and 12.550 respectively,the bidding times of 82% of the nodes are normal.The weighted degree value of 2% nodes seriously exceeds the average value,and the bidding times of these nodes were abnormal.The average degree value and average weighted degree value of bidder network of supervision project are 11.631 and 26.306 respectively.It is higher than the level of construction projects,survey and design projects and the overall network,and the bidding behavior is the most active.(3)The degree of collusion suspicion of each node in a typical community is analyzed.According to the index of overlap and clustering coefficient of nodes,3144 and 2716 nodes in 128 construction associations belong to overlap structure,16 enterprises have high degree of suspicion,10 enterprises have low degree of suspicion,and 54 enterprises have very low degree of suspicion.(4)It is verified that the complex network method is helpful to identify the collusion behavior of bidders.The bidding price of twelve of the bidding projects jointly participated by Nos.53,68 and 1251 identified by complex network technology has obvious anomaly regularity.According to the probability of winning the bid,it can be found that most of the nodes in the community achieve the goal of winning the bid through collusion.In addition,by comparing the suspected collusive bidding enterprises published in S Province in 2017,it is found that some bidders are divided into the same community according to historical records.It proves the validity and feasibility of using complex network method to identify conspiracy behavior.Based on the historical data of successful bidding for construction projects,this study uses complex network technology to identify covert conspiracy behavior.The research results broaden the thinking of the study of conspiracy behavior for construction projects.At the same time,it provides a reference for government supervision departments to promote the wisdom of supervision in the future with the help of electronic bidding and trading data.