Structured simulation data-oriented quality evaluation method, system and equipment

Abstract

The invention provides a structured simulation data-oriented quality evaluation method and system. The method comprises the steps of constructing a simulation data generation model according to original data; the data owner uses original data containing privacy information to train a simulation data generation model to obtain simulation data; the original data and the simulation data are input into an evaluation system, the availability and privacy of the simulation data are evaluated, and the simulation data with the highest quality is selected to be published; the original data and the simulation data are structured data and are composed of a plurality of data records, each record has a plurality of fields, and the fields of the simulation data are completely the same as the fields of the original data. According to the method, the simulation data is used for replacing the original data to participate in data circulation for quality evaluation, the data items of the simulation data are synthesized and are not associated with any entity in reality, and other hypotheses or priori knowledge are not needed, so that the risk of privacy disclosure is effectively avoided, the evaluation is comprehensive and reliable, and the universality is good.

Description

Technical field [0001] The present invention relates to the field of data security technology, and more particularly to an adaptation method, apparatus, and storage medium facing a watermark for relational database. Background technique [0002] Big Data Age, data owner and collaborators share data or externally released data. Traditional data anonymization technology cannot ideally protect privacy. For attackers with relevant other data source knowledge, it is likely to speculate that the anonymous processing identifier or ratio identifier, that is, the recognition attack; and the data is anonymous The technology will significantly reduce the availability of data. [0003] Patent Document CN110222058A, disclosed in China's National Intellectual Property Office, proposes a multi-source data associated privacy lepse risk assessment system based on FP-GROWTH, which needs to be associated with external related data sources, analyzing results and collected external data source height...

AI Technical Summary

Problems solved by technology

[0004] In the above-mentioned published literature, a large amount of relevant prior knowledge of the data set is required, such as the privacy attribute of the data set, the analysis task goal, the associated data set, etc.; it can only be evaluated for the data set after desensitization or anonymization. It is not suitable for evaluating simulation data; the data set can only be evaluated from one aspect of privacy, and the evaluation angle is not enough

Images