Research on accurate perception method of multi-source and multi-modal data based on data features

In the power communication system, highly dynamic and real-time data acquisition often leads to incompleteness, error information, and vulnerability. In order to solve this problem, a data privacy protection method for power communication system based on differential privacy was proposed. By classifying the different attributes of the original dataset, the feature set and candidate set are constructed. Then, the feature classification model based on the maximum information coefficient was used to analyze the two datasets, and the data with the highest correlation were extracted from them to form a privacy dataset. Using differential privacy technology, a protected anonymized dataset is generated through a data anonymization privacy protection model to ensure data privacy security. The test results show that the model has excellent computing performance under the premise of maintaining a reasonable privacy protection budget, and the connection value of the processed data record is reduced to less than 0.23.