Contractor Recommendation Based On Credit Network

Owners and contractors are the main players in the construction market.Whether the owner can choose the right contractor among the large number of contractors is related to the success or failure of the project.With the information sharing of the construction industry and the construction of the credit platform,the owner can evaluate the contractor from the perspective of credibility,to avoid the risk of trust loss caused by industry competition and moral deficiencies.The source of the contractor’s credibility is the evaluation and recommendation of the contractor by the previous owner who has cooperated with the contractor.When using these information,the decision-making owner will fully consider the relationship between himself and the former owner.Only the recommendation from the owner who is trusted by the decision-making owner can be used for reference so as to find the contractor who meets his own expectations.In this way,the efficiency of the decision-making owner is improved,which contributes to the benign development of the overall credit of the construction market.This paper draws on the user-based collaborative filtering algorithm of the computer,proposing a credit network recommendation model based on the owner-based collaborative filtering algorithm,to solve the trust of the decision making owner to the former owner and obtain valuable contractor recommendation information.The model has experienced stages of related information collection,neighbor set formation,contractor credit evaluation and recommendation list formation.The important content of the model is to form the neighbor set of the owner.The comprehensive trust of the decision-making owner to the former owner is composed of the composite trust value and the similarity between them.The former owner whose comprehensive trust is higher than a certain standard forms the neighbor set of the decision-making owner.The composite trust value consists of direct or indirect trust values and credit values.The former considers the communication and cooperation between the owners,while the latter considers the comprehensive evaluation of the construction market to the owners.Trust is transmitted by trust chain.Credit propagation is similar to the principle of page sorting by Page Rank algorithm,and the credit value is calculated according to its rules.The similarity consists of the similarity between the owner’s trust evaluation for contractors and common attributes of owners.Since the owner’s cooperation time will affect the trust relationship between the owners,the time decay function is introduced to correct the difference in trust caused by time.The recommendation from neighboring owners makes the decision-owner more satisfied.Calculate the credit score of the contractor to be cooperated,and the contractor with the higher score can enter the recommendation list.Simulate a real case and use simulation design to observe the ranking.The results show that the model is effective.This paper focuses on the trust relationship between owners and the credit evaluation of multiple subjects on the same goal.With reference to computer algorithms,the credit ranking of contractors has been achieved,and a list of candidate contractors with high credit has been obtained,which provides a new effective method for owners to choose contractors.