Research And Implementation Of Manufacturing Task Data Release Scheme Based On Safe K-anonymity For Privacy Protection

Cloud manufacturing is a new type of manufacturing mode that achieves the high sharing of social resources.With advanced information technology,cloud manufacturing integrates resources shared by resource owners in the manufacturing life cycle activities,and provides standardized and shared manufacturing services for different regional cloud users.But with the proliferation of data sources to cloud manufacturing platforms and data mining technologies,the release of these production data can reveal important information about the business.Therefore,it is of great significance to study data publishing technologies that can balance the security and effectiveness of information with the cloud resource sharing.In this paper,the manufacturing task data produced in the manufacturing process of the enterprise is taken as the object.According to the problem of disclosure of private information of the data release,this paper conducts an in-depth research on the key technologies involved in the data security release and provides technical support for sharing data resources.The main research contents are as follows:(1)Based on the existing problems in manufacturing resource sharing,this paper compares and studies the existing data security publishing technologies,the characteristics and leakage cases of the manufacturing task data through analysis of the security issues after the release of manufacturing task data,and builds the safe K-anonymity data security release framework for manufacturing tasks,and discusses the key technologies involved in the framework in-depth study.(2)In view of the existing classification of safe K-anonymity and the non-uniform standard of data generalization,combining with the characteristics of each attribute in manufacturing data,this paper studies the method of identifying and classifying attributes,optimizes the anonymous generalization of the original dataset based on different attribute information and top-down local generalization algorithm.In order to avoid the disclosure of the relevancy in the data after the data is the depth of mining,the data relevance needs to be evaluated and processed after the data anonymization.According to the data owner's posting threshold of data relevance,a low relevancy packet switched publish algorithm of data record is proposed.(3)Designing and implementing a data security publishing system for privacy protection based on the actual manufacturing task data publishing scenario.The system simplifies the process of data release and meets the data publishers' protection of privacy information and the user's need for data and information.The system can provide users with functions,including classification of data attribute information identification,data generalization,low correlation processing and comparative analysis of published data.Testing by real manufacturing task data in publishing system,data security release method of manufacturing tasks that we study and implementation of can retain more information,in the controllable range of time and information safe.