Title:

Cleansing Indoor RFID Tracking Data

Abstract:

RFID is increasingly being deployed for object tracking in indoor settings. For example, many airports in Scandinavian countries employ RFID for monitoring and tracking bags in their baggage handling systems. However, raw indoor RFID tracking data thus obtained is characterized by noise (cross readings) and incompleteness (missing readings), which should be cleansed before the data is used meaningfully at higher levels. This talk introduces two approaches to indoor RFID data cleansing. The graph based approach captures the detailed knowledge about the indoor setting and RFID deployment in an information-rich graph that in turn facilitates data cleansing. In contrast, the learning-based approach requires considerably less prior knowledge but adapts a Markov model to capture the uncertainty in indoor RFID data. By learning the model parameters from enough raw data, this approach achieves a data cleansing accuracy comparable to or even better than what the graph-based approach can deliver.

Bio:

Hua Lu is an associate professor in the Department of Computer Science, Aalborg University, Denmark. He received the BSc and MSc degrees from Peking University, China, and the PhD degree in computer science from National University of Singapore. His research interests include database and data management, geographic information systems, and mobile computing. Recently, he has been working on indoor space data management, skyline queries, complex spatial queries, and social media data management. He has served on the program committees for conferences and workshops including ICDE, CIKM, ACM SIGSPATIAL, SSTD, MDM, PAKDD, APWeb, WAIM, and MobiDE. He has also served as PC cochair or vice chair for ISA 2011, MUE 2011 and MDM 2012, demo chair for SSDBM 2014, and PhD forum cochair for MDM 2016. He is a senior member of the IEEE.

« Go Back