摘要
针对目前云存储环境下主流的SQLite数据库系统存在各类安全问题而缺乏数字取证方式的情况,提出了一种对SQLite数据库文件进行数据恢复来取证的方法。针对现有的数据恢复方法无法准确定位数据块头部字段,恢复范围小,恢复准确率低的问题,通过SVM分类器训练线性归一化模型,定位待恢复区域各单元块起始字段,分离各数据字段并利用特征匹配的方法对SQLite数据库文件进行数据恢复,具有高适用范围并有效提高了恢复准确率。
With the application and the popularization of SQLite database, numerous of intelligent termi- nal applications and cloud storage devices use SQLite database as a data storage mode. According to the existing security problems of the mainstream SQLite database system in the cloud storage environment and lack of digital forensics, a new data recovery scheme for SQLite database files is proposed. In the existing data recovery schemes, the head field of the data block cannot be accurately located, the recovery range is smaller and the recovery accuracy is lower. To solve the above problems, the proposed scheme for the recovery of SQLite data files uses the SVM classifier training and the linear normalization model to restore regional localization start field and separate data fields. Based on the model building, the feature matc- hing scheme is used to restore the data of the SQLite database file. The scheme can provide a high appli- cability, thus imDrovinz the accuracy of the recovery rate.
出处
《南京邮电大学学报（自然科学版）》
北大核心
2018年第1期106-112,共7页
Journal of Nanjing University of Posts and Telecommunications：Natural Science Edition
基金
国家自然科学基金(61372103)
信息网络安全公安部重点实验室开放课题(C16602)资助项目
