Marine Big Data Sensitivity Assessment System and Prevention Methods for Confidentiality Needs

Abstract

The invention discloses a marine big data sensitivity evaluation system and a marine big data sensitivity prevention method for confidentiality requirements. The marine big data sensitivity evaluationmethod comprises the following steps: step 1, establishing a sensitive feature library; Step 2, matching and analyzing a to-be-processed data set; 3, constructing a sensitivity assessment model; 4, calculating attribute sensitivity; 5, protecting sensitive data; And 6, dynamically optimizing the sensitive feature library. The method is used for realizing identification and protection of sea-related data with different data sensitivities and value intensities from various marine service systems.

Description

technical field [0001] The invention belongs to the technical field of data identification and processing, and in particular relates to a marine big data sensitivity assessment system and prevention method for confidentiality requirements. Background technique [0002] Data is an important asset of organizations, enterprises and individuals, and data leakage will cause great losses to data owners. In recent years, data leakage has become the source of the black industry chain of information trafficking. The security protection of sensitive data in the big data environment has extremely high theoretical research value and engineering practice significance. How to protect sensitive data has become the focus of big data security. [0003] Existing technologies mostly adopt: access control, sensitive data confidentiality technology based on data distortion (retaining certain statistical characteristics), data encryption, and restricted release (not releasing certain attributes ...

AI Technical Summary

Problems solved by technology

[0007] The sensitive data dictionary matching method has the following defects: 1. The recognition accuracy is low. The dictionary matching adopts a pattern matching method, so the establishment of the data dictionary determines the accuracy of sensitive data recognition. When the dictionary is incomplete or the dictionary is established incorrectly , there will be a problem of reduced recognition accuracy; 2. Classification result interference. Because dictionary matching is used, the same data information will be matched to multiple data dictionaries. Because traditional data dictionaries cannot perform weighted calculations, it will cause interference to classification results. , resulting in inaccurate classification results

[0008] The artificial identification method of sensitive data has the following defects: 1. The identification speed is slow. Due to the manual processing method, when faced with a large amount of data, the manual combing speed is relatively long compared with the machine identification speed cycle, and the professional quality of the processing personnel is relatively high. High; 2. The judging standards are not uniform. Since the sensitive data identification process mainly depends on the subjective judgment of people, different people may have different judging standards for the same data, and even the results identified by the same person at different times are still different. different, which will lead to discrepancies in sensitive data identification results

The self-built corpus may have the problem that the corpus is not comprehensive enough due to the insufficient amount of training data

In step 2, it is still necessary to manually identify and classify sensitive features of the basic corpus, which is highly subjective. Different people may have different criteria for judging the same data, and even the same person may identify the same data at different times. The results are still different, which will lead to discrepancies in the sensitive data identification results

In step 3, the category of the target data is obtained through the weighted ranking of sensitive words, but the sensitivity of the data is not quantified

Images