CAJ | 학술논문

随着大数据环境下数据安全问题逐渐增加，潜在破坏信息量也不断增加，传统应用决策树的信息分层隔离方法，处理的破坏信息量存在一定的限制，无法实现大规模破坏信息的准确分层隔离。提出一种应用动态分层和最优熵滤波器的破坏信息分层隔离方法，在大数据环境下破坏信息的静态分层的基础上，分析了潜在破坏信息的动态分层，塑造破坏信息状态模型与破坏信息隔离模型，通过滤波器对破坏信息的状态进行预测，将潜在破坏信息隔离问题简化成熵最优化问题，获取破坏信息状态误差的概率密度函数，确保破坏信息状态误差的概率密度函数的熵最小化，分离出潜在破坏信息。仿真实验结果说明，所提方法在不同破坏信息数量条件下，对潜在破坏信息的分层隔离误差低于传统方法，并且具有较高的分层隔离效率。
수착대수거배경하수거안전문제축점증가，잠재파배신식량야불단증가，전통응용결책수적신식분층격리방법，처리적파배신식량존재일정적한제，무법실현대규모파배신식적준학분층격리。제출일충응용동태분층화최우적려파기적파배신식분층격리방법，재대수거배경하파배신식적정태분층적기출상，분석료잠재파배신식적동태분층，소조파배신식상태모형여파배신식격리모형，통과려파기대파배신식적상태진행예측，장잠재파배신식격리문제간화성적최우화문제，획취파배신식상태오차적개솔밀도함수，학보파배신식상태오차적개솔밀도함수적적최소화，분리출잠재파배신식。방진실험결과설명，소제방법재불동파배신식수량조건하，대잠재파배신식적분층격리오차저우전통방법，병차구유교고적분층격리효솔。
As the data security problem increasingly large data environment, potential damage is also increasing amount of information, the traditional application of decision tree information layered isolation method, there are certain restrictions, dealing with the destruction of the information cannot be achieved massive destruction information accurately layered isolation. Put forward a kind of the destruction of the application of dynamic hierarchical and optimal entropy filter information layered isolation method, in the big data environment damage information of static stratification, on the basis of analysis of the potential damage information dynamic hierarchical, mold damage information state model and damage information isolation model, through the filter, to predict the damage state of the information to simplify the potentially damaging information isolation problems ChengShang optimization problem, the damage information state error probability density function, to ensure that the damage information state error probability density function of the entropy minimization, isolate potentially damaging information. The result of simulation experiment shows that the proposed method under the condition of different damage information quantity, layered isolation of potential damage information error is lower than the traditional method, and has high layered isolation efficiency.