Constrained min-cut replication for K-way hypergraph partitioning

Date

2014

Editor(s)

Advisor

Supervisor

Co-Advisor

Co-Supervisor

Instructor

BUIR Usage Stats
4
views
32
downloads

Citation Stats

Series

Abstract

Replication is a widely-used technique in information retrieval and database systems for providing fault tolerance and reducing parallelization and processing costs. Combinatorial models based on hypergraph partitioning are proposed for various problems arising in information retrieval and database systems. We consider the possibility of using vertex replication to improve the quality of hypergraph partitioning. In this study, we focus on the constrained min-cut replication (CMCR) problem, where we are initially given a maximum replication capacity and a K-way hypergraph partition with an initial imbalance ratio. The objective in the CMCR problem is finding the optimal vertex replication sets for each part of the given partition such that the initial cut size of the partition is minimized, where the initial imbalance is either preserved or reduced under the given replication capacity constraint. In this study, we present a complexity analysis of the CMCR problem and propose a model based on a unique blend of coarsening and integer linear programming (ILP) schemes. This coarsening algorithm is derived from a novel utilization of the Dulmage-Mendelsohn decomposition. Experiments show that the ILP formulation coupled with the Dulmage-Mendelsohn decomposition-based coarsening provides high quality results in practical execution times for reducing the cut size of a given K-way hypergraph partition. © 2014 INFORMS.

Source Title

INFORMS Journal on Computing: charting new directions in OR and CS

Publisher

Institute for Operations Research and the Management Sciences (I N F O R M S)

Course

Other identifiers

Book Title

Degree Discipline

Degree Level

Degree Name

Citation

Published Version (Please cite this version)

Language

English