Repair Singleton IDs on the Fly

摘要：

　　Tracking moving entities at predefined locations plays an essential role in many surveillance related applications.Occasionally,the IDs of those entities are incorrectly recorded due to various reasons such as errors in recognition.Such errors need to be repaired on the fly as those IDs are often involved in some time-sensitive query processing or data analysis tasks.In this paper,we address a specific case where the errors result in singleton IDs,i.e.,IDs that appear only once during a specific period of time and thus could be safely presumed to be erroneous.The repair of the IDs is based on constraints posed by the data itself (e.g.,constraints posed by the road network).We present a tracking tree structure to index the candidate repairs for each singleton ID,which enables repairing of the IDs on the fly.We implement a distributed repair system on the Apache Storm platform.Experiments on both real and synthetic datasets demonstrate the effectiveness and efficiency of our singleton detection and repair approach.

关键词： Data quality Location sequence constraint Stream process

作者: Xingcan Cui De Guo De Guo

作者单位: School of Computer Science and Technology,Shandong University,Jinan,China School of Computer Science and Technology,Shandong University,Jinan,China;School of Information Tech

会议类型: 国际会议

会议名称: International Asia-Pacific Web Conference(第18届国际亚太互联网大会)

会议地点: 苏州

会议语种:英文

页码: 214-226

在线出版日期: 2016-09-23（万方平台首次上网日期，不代表论文的发表时间）