Graph Mining for Big Data Problems using MapReduce
发布人: 曹思圆   发布时间: 2018-07-05   浏览次数: 10

主讲人:吕恩月 教授
Department of Mathematics & Computer Science at Salisbury University,USA
主持人:吕长虹 教授
开始时间:2018-7-7 上午10:30-11:30
主办单位:数学科学学院 科技处

报告人简介:Dr. Enyue Lu received her Ph.D. in Computer Science from University of Texas at Dallas in 2004. Currently, she is a full professor of computer science in the Department of Mathematics & Computer Science at Salisbury University and director and PI of NSF funded project: “REU SITE: EXERCISE - Explore Emerging Computing in Science and Engineering.” Her research interests include high performance computing, computer networks and security, parallel and distributed algorithm design and analysis, and graph theory. She has published over 50 peer-reviewed papers, supervised a total of 46 students in 30 different undergraduate research projects and internships, obtained multiple undergraduate research grants of over one million dollars from NSF, and served at over 30 school and professional committees.

Graph mining for big data problems has many applications including community identification, blog analysis, and intrusion and spamming detections. Currently, it is impossible to process real-world big data problems using a single computer. In this talk, we take advantage of MapReduce, a programming model for processing large data sets, to detect important graph patterns using open source Hadoop on Amazon EC2. The aim of this talk is to show how MapReduce cloud computing with the application of graph pattern detection scales on real world data.  We will show MapReduce graph algorithms to enumerate graph patterns including triangles, trusses, and barycentric clusters. The performance comparison for MapReduce graph algorithms using real world data taken from Snap Stanford will be discussed too.  We’ll also show how graph clustering can be used to solve real-world network intrusion detection problems.