首页> 外文会议>IEEE International Conference on Data Mining Workshops >BeTracker: A System for Finding Behavioral Patterns from Contextual Sensor and Social Data
【24h】

BeTracker: A System for Finding Behavioral Patterns from Contextual Sensor and Social Data

机译:Betracker:用于查找来自上下文传感器和社交数据的行为模式的系统

获取原文

摘要

In this work, we integrate the contextual information provided from sensor data and the social relationships collected from online social networks to construct a system, termed Be Tracker. We aim to find and track the frequent and representative behaviors for any user-input individual or social structural information. We claim combining physical contacts from sensor data and virtual online interactions can reveal real-life human behaviors. In our Be Tracker, we mine the temporal sub graph patterns as the discovered behaviors from sensor-social data transactions. The user-given information, which is the target to observe, can be (a) an individual (to find her daily behaviors), (b) a relational structure (e.g. linear, triangle, or star structure) (to find the frequent and contextual interactions between them), and (c) a relational structure with partially assigned individuals and sequential time periods (to observe their interactions that follow certain temporal order). In the experimental part, we demonstrate promising results of different queries and present the system efficiency of the proposed behavioural pattern mining.
机译:在这项工作中,我们集成了传感器数据提供的上下文信息,以及从在线社交网络收集的社交关系来构建系统,称为跟踪器。我们的目标是找到和跟踪任何用户输入个人或社会结构信息的频繁和代表性。我们要求从传感器数据和虚拟在线互动组合物理接触可以揭示现实生活的人类行为。在我们的追踪器中,我们将时间子图模式作为来自传感器社交数据事务的发现行为。用户给定的信息是观察的目标,可以是(a)个人(找到她的日常行为),(b)关系结构(例如线性,三角形或星形结构)(用于找到频繁和它们之间的上下文相互作用),(c)具有部分分配的个体和顺序时间段的关系结构(观察其遵循某些时间顺序的相互作用)。在实验部分中,我们展示了不同疑问的有希望的结果,并提出了所提出的行为模式挖掘的系统效率。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号