Information Security of Marketing Network Platforms in Big Data Era: Consumer Data Protection

Abstract

In traditional marketing network platforms, consumer data is collected and analyzed on a large scale and used for precision marketing. Traditional encryption and access control methods cannot guarantee the real-time and security of data, and there is a risk of privacy leakage in cross-platform data exchange. This paper builds a cross-platform data protection framework, uses the Paillier encryption algorithm to encrypt consumer data, and ensures that additive aggregation calculations can be performed directly in the encrypted state. This paper introduces Laplace distribution noise into the data statistical results, and balances data practicality and privacy protection by dynamically adjusting the privacy budget; attribute-based access control (ABAC) is adopted. Permissions are dynamically assigned based on user roles and environment variables, and the permission management tool updates access rules in real time to ensure that sensitive data is only accessible to legally authorized users. The LSTM (Long Short-Term Memory) time series analysis model can be used to learn access logs, establish a normal behavior baseline, detect abnormal data access behavior in real time, and use the early warning system to prevent potential threats. Experimental results show that under the data volume of 10,000, the entropy value of the Paillier encryption algorithm reaches 0.83, the average response time is 7.0 seconds, and the computing throughput is 1680 KB/s, which has higher data real-time and security. Under the 100% noise ratio of reconstruction attack, camouflage attack, reverse engineering attack and association attack, the signal-to-noise ratio (SNR) of the Laplace mechanism is 0.74, 0.73, 0.71 and 0.73 respectively, which has a high privacy protection effect. The experiment proves the effectiveness of this paper's research on consumer data protection.