大数据下移动终端应用的用户隐私泄露分析模型方法研究

Privacy leak problem has become one of the most serious problems for cyber security. Lots of security events involving a large number of users bring adverse social impact. Privacy leak problem disturbs the network life of users and impairs the development of Internet. It has become an urgent problem for Internet security. Analysis of privacy leak has become a significant research direction. Therefore, we perform an in-depth discussion on privacy leak issues, study the modeling and analysis method of privacy leak software behavior. .Modeling and analysis of malware behavior is an important research issues of wide concern in recent years. There are many mature theoretical models and analysis methods. But several problems arise when focusing on the privacy leak software behavior. (1) Existing theoretical models often used more general model for malware analysis. But the existing models were not effective roles for privacy leak software behavior; (2) Existing works for privacy leak issues gave only a relatively simple analysis, lacked a complete analysis architecture; (3) Existing works were for a unified analysis of malware, complicated code obfuscating the transforming technology often made these methods powerless..Therefore, we present a new modeling and analysis approach to solve the privacy leak software behavior analysis problem. Our main research content includes the following parts..1..STPPN model for privacy leak software behavior of mobile apps with big data environment.2..Qualitative and quantitative analysis based on STPPN model.3..In-depth analysis for multiple device collaboration behavior .4..Design and Implementation of prototype system.In conclusion, our graphical, modularized and computational PPN can model and compute privacy leak software behavior effectively and precisely. The analysis methods based on PPN model can provide comprehensive results from multiple aspects of privacy leak software behavior. Complicated behavior such as multiple process collaboration behavior and high crypticity behavior can also be analyzed in depth. The experimental results on real world software show that our approach privodes solid theoretical support to privacy leak software behavior analysis by comprehensive and accurate output. Our prototype system is also a useful tool for actual practice.

隐私泄露问题极大的影响了用户的正常网络生活 ，相关研究已成为具有重要意义的研究方向。本项目提出针对大数据下移动应用隐私泄露行为的STPPN模型，以及基于模型的单终端动态演化分析方法与多终端关联分析方法，并实现相应的原型系统。本项目利用STPPN独有的“时”“空”建模能力，对于时空变化的用户隐私泄露情况进行演化分析，解决大数据下移动终端应用行为的动态非入侵式的模式提取问题；对于多终端上的移动应用的用户隐私泄露情况进行互动式行为染色分析和共性行为群体分析，解决多应用行为的关联分析问题；利用STPPN的模块化和层级化特性，解决大数据量下的应用行为的状态空间规模爆炸问题；利用STPPN特有的“标识可达性”（token reachability）提供数据流追踪能力，解决大数据下的大规模用户隐私泄露情况难以追踪和溯源的问题。通过上述研究为我国互联网隐私保护提供理论基础和实用工具。

隐私泄露问题极大的影响了用户的正常网络生活 ，相关研究已成为具有重要意义的研究方向。本项目提出针对大数据下移动应用隐私泄露行为的STPPN模型，以及基于模型的单终端动态演化分析方法与多终端关联分析方法，并实现相应的原型系统。本项目利用STPPN独有的“时”“空”建模能力，对于时空变化的用户隐私泄露情况进行演化分析，解决大数据下移动终端应用行为的动态非入侵式的模式提取问题；对于多终端上的移动应用的用户隐私泄露情况进行互动式行为染色分析和共性行为群体分析，解决多应用行为的关联分析问题；利用STPPN的模块化和层级化特性，解决大数据量下的应用行为的状态空间规模爆炸问题；利用STPPN特有的“标识可达性”（token reachability）提供数据流追踪能力，解决大数据下的大规模用户隐私泄露情况难以追踪和溯源的问题。在国内外本领域重要的学术刊物和会议上发表论10篇，其中本领域SCI 期刊1篇，在其他国际学术会议和国内一级学报上发表EI检索论文9篇，其中CCF-C类3篇，部分论文在国际上有一定的引用率；申请国家发明专利或软件著作权2项；出版相关专著1部；设计并实现网络消费群体行为分析试验验证（原型）系统；.