Извините, регистрация закрыта. Возможно, на событие уже зарегистрировалось слишком много человек, либо истек срок регистрации. Подробности Вы можете узнать у организаторов события.
Mining frequent substructures from graphs or networks is important to many applications, such as bioinformatics, social network analysis, software engineering, and recommendations. The existing studies are mostly designed for mining frequent substructures from a set of small graphs. However, nowadays, we often have data sets modeled by a very large single graph, e.g. a social network. This makes the task even hard because of complex overlaps of substructures.
Mining frequent substructures from graphs or networks is important to many applications, such as bioinformatics, social network analysis, software engineering, and recommendations. The existing studies are mostly designed for mining frequent substructures from a set of small graphs. However, nowadays, we often have data sets modeled by a very large single graph, e.g. a social network. This makes the task even hard because of complex overlaps of substructures. The growing sizes of graphs as well as the interesting frequent substructures make the task even challenging. To discovery a large substructure, we suppose to examine all the small substructures that could lead to the large one, which is exponential and expensive to the graph sizes. In this talk, Qiang is going to analyze the problem and introduce a new graph mining tool "Spider Framework" to solve the new challenges. Specifically, this method leverages the properties of small frequent structures to efficiently find large frequent substructures. This method can speed up to a factor of 10 on a random graph comparing with one of the most efficient method SUBDUE that randomly reports small frequent substructures.
Qiang Qu is an assistant professor at Innopolis University. He obtained his Ph.D. in Computer Science from Aarhus University, supported by the GEOCrowd project under Marie Sklodowska-Curie Actions, his M.Sc in Computer Science from Peking University, and B.Sc. in Management Science from Dalian University of Technology. He was a research scholar to ETH Zurich, Carnegie Mellon University, Singapore Management University, and Nanyang Technological University. His current research interests are in spatial-rich data management, querying processing, large-scale data mining, and cognitive computing.