Password attribute analysis method based on neural network

Abstract

The invention discloses a method for analyzing password data in the field of computer security, and particularly relates to a system for analyzing the attribute relation of real password data based on neural network modeling to obtain valuable information. The method comprises the following main steps: acquiring a password attribute data source, determining a neural network model, carrying out training learning, carrying out a simulation test and the like. The password attribute data source is obtained by endowing each password with a group of attributes and performing assignment according to password compositional features in sequence; the neural network model comprises a network type, input and output attributes, the number of network layers, the number of neutrons of each layer, training sample and simulation sample capacity, and relevant function and parameter setting; the training learning is to obtain a relation model closest to a real one by continuously correcting the model according to a certain quantity of attribute data training samples; the simulation test is to test a trained network according to a certain quantity of attribute data test samples; whether a valuable relation and a rule exist between the input attribute and the output attribute or not is judged according to the simulation accuracy, thus instructing a dictionary file to be generated and providing directive assistance for cracking and recovery of the password.

Description

technical field [0001] The invention relates to an analysis method for password data in the field of computer security, in particular to a system for analyzing and processing attribute relationships of real password data based on neural network modeling to obtain valuable information. Background technique [0002] Artificial neural network (ANN for short), is a complex network system interconnected by basic processing units called neurons, referred to as neural network (NN for short). A mathematical or computational model of the structure and function of a network. Neural networks are computed by a large number of artificial neuron connections. In most cases, the artificial neural network can change the internal structure on the basis of external information, which is an adaptive system. A modern neural network is a nonlinear statistical data modeling tool that is often used to model complex relationships between inputs and outputs, or to explore patterns in data. [0003...

AI Technical Summary

Problems solved by technology

[0007] In further novelty searches, no literature reports identical or similar to the subject of the present invention have been found

Images