基于分布式序列模式挖掘物联网服务用户隐私保护关键技术研究

项目名称： 基于分布式序列模式挖掘物联网服务用户隐私保护关键技术研究

项目编号： No.61303193

项目类型： 青年科学基金项目

立项/批准年度： 2014

项目学科： 自动化技术、计算机技术

项目作者： 冯云霞

作者单位： 青岛科技大学

项目金额： 23万元

中文摘要： 针对物联网及开放的云计算环境，采用分布式序列模式挖掘技术建立用户群体行为模式。基于该模式开发的个性化物联网服务不仅能更好地满足用户需求，而且能提高效率，从而节省资源。但是，分布式序列模式挖掘也使用户面临比互联网服务系统更严重的隐私安全威胁。本项目从技术角度研究基于分布式序列模式挖掘物联网服务用户敏感数据和敏感序列模式的保护问题。首先，以位置和运动轨迹为例，研究了用户敏感数据保护中的关键科学问题，分别提出基于混合策略的位置安全定位技术和基于多元回归分析的移动轨迹保护机制。重点研究分布式序列模式挖掘用户的敏感序列模式保护问题，提出基于同态加密算法和多方安全计算的隐私保护算法和协议。本项目在位置隐匿、轨迹隐藏和用户敏感序列模式保护等方面取得的研究成果可方便地扩展到其他隐私保护领域。本项目可提升我国物联网服务研究和应用的技术水平，为构建高效而安全的物联网服务提供有价值的参考。

中文关键词： 物联网用户隐私保护；基于移动终端的隐私保护；位置-查询隐私保护；异构大数据集成框架；分布式时序大数据挖掘

英文摘要： Under cloud computing environment of IOT, it is appropriate to adopt distributed sequential pattern mining to establish the behavior pattern of users. Based on which, personal IOT Services may not only better meet the user needs, but also save resources dramatically. Nevertheless, distributed sequential pattern mining aggravates the problem of privacy threats. This project researches problems of preserving both privacy data and sensitive sequential pattern of users in distributed sequential pattern mining based IOT service system. First, we research public scientific problems of data privacy preserving taking position and trajectory as examples, and put forward mixed location privacy preserving position technology and a multiple regression analysis based mobile trajectory protection scheme, respectively. We focus our attention on protection of sensitive sequential pattern during distributed sequential pattern mining. Achievements we achieved on location anonymity, trajectory preserving and sensitive sequential pattern protection can conveniently extend to other privacy preserving fields. This project may greatly promote the theory level of both IOT services and applications technology. Furthermore, it provides valuable references to establish efficient and safe IOT services.

英文关键词： Privacy protection for IOT users；Mobile terminal based location privacy protection；location-query privacy protection；Integration architecture of heterogeneous big data；Distributed sequential big data mining