Classification of Encrypted Text Based on Artificial Intelligence
Bo Zhang · IOP Conference Series Materials Science and Engineering · 2020
Abstract With the development of internet, information security is essential. The traditional decryption method only works for single encryption method. New models based on Artificial Intelligence to decrypt different kinds of encryption methods. Data features were extracted by tf-idf algorithm and applied into different models. The primary models to decrypt message without knowing whether encryption methods are logistic regression, lightGBM, and ensemble algorithm to decrypt encryption data. Comparing the performance in the models based on logistic regression and lightGBM algorithm respectively, we concluded that both the accuracy and F1 score of logistic algorithm are better than those of lightGBM algorithm. We also made improvement on the model by using ensemble algorithm. A more preferable performance including the accuracy and F1 score has achieved with multiple algorithm. Therefore, ensemble algorithm is more suitable for building general model to decrypting different kinds of encryption methods.