Trajectory Data Mining

alt
  • Jae-Gil Lee, Ph.D
  • Associate Professor
  • Department of Industrial & Systems Engineering
  • Korea Advanced Institute of Science and Technology (KAIST)
  • Korea

Abstract
 The widespread use of location-aware mobile devices has made it feasible to collect precise locations of the users. Additionally, traditional social networking services such as Foursquare and Twitter have adopted geo-tagging capabilities to capture the location preferences of their users. Along these lines, a majority of location-based services of today retain abundant data about users’ location preferences (e.g., travel routes or trajectories). In this tutorial, I explain popular trajectory data mining techniques in two aspects. In the first part, I will deal with various definitions of group movement patterns, such as flock patterns, trajectory patterns, convoy patterns, and swarm patterns, and the algorithms for finding those patterns. In the second part, I will discuss recent trends that consider user locations and social relationships all together to improve the quality of knowledge discovered, focusing on various geosocial group querying suggested for geosocial networks.

Biography
 Jae-Gil Lee is an Associate Professor at Graduate School of Knowledge Service Engineering, KAIST. His research interests encompass spatio-temporal data mining, social-network and graph data mining, and big data analysis with Hadoop/MapReduce and Spark. His work on trajectory clustering published at ACM SIGMOD 2007 is the most cited paper in trajectory data mining. In addition, he received the Best Paper Award at AAAI ICWSM 2013 and recently served as the Program Committee Co-Chair of PAKDD 2017.
ⓒ Copyright 2017 KIISE – All Rights Reserved.
[KIISE] Korean Institute of Information Scientists and Engineers
#401 Meorijae Bldg., 76, Bangbae-ro, Seocho-gu, Seoul 137-849, Korea http://www.kiise.or.kr
+82-2-588-9240 / +82-2-521-1352 / mkim@kiise.or.kr
Business Registration Number: 114-82-03170