Heavyweight Pattern Mining in Attributed Flow Graphs

• Author / Creator

  Simoes Gomes, Carolina

• Flow graphs are an abstraction used to represent elements travelling through a network of nodes. The paths between nodes are directed edges in the graph, and the amount or transmission frequency of elements that go through the paths are edge weights. If additional data is associated with the nodes, we have attributed flow graphs (AFGs). This thesis defines heavyweight patterns, which are sub-sets of attributes connected by edges found in a dataset of AFGs, and have a computed weight higher than an user-defined threshold. The thesis also defines Heavyweight Pattern Mining, the problem of finding heavyweight patterns in AFGs. It presents a new algorithm called AFGMiner, which solves Heavyweight Pattern Mining and associates patterns with their occurrences in the dataset. In
  addition, it describes HEPMiner and SCPMiner, two new program performance analysis tools that apply AFGMiner and have as target users compiler and application developers respectively.

• Subjects / Keywords

  • Sub-graph mining
  • Data mining
  • Program profiling
  • Flow graph
  • Program analysis
  • Software analysis
  • Pattern mining

• Graduation date

  Fall 2012

• Type of Item

  Thesis

• Degree

  Master of Science

• DOI

  https://doi.org/10.7939/R3HW5C

• License

  This thesis is made available by the University of Alberta Libraries with permission of the copyright owner solely for non-commercial purposes. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.