Abstract:Balanced partitioning is often a crucial first step in solving large-scale graph optimization problems, e.g., in some cases, a big graph can be chopped into pieces that fit on one machine to be processed independently before stitching the results together. In other cases, links between different parts may show up in the running time and/or network communications cost. We study a distributed balanced partitioning problem where the goal is to partition the vertices of a given graph into k pieces so as to minimize the total cut size. Our algorithm is composed of a few steps that are easily implementable in distributed computation frameworks. The algorithm first embeds nodes of the graph onto a line, and then processes nodes in a distributed manner guided by the linear embedding order. We examine various ways to find the first embedding, e.g., via a hierarchical clustering or Hilbert curves. Then we apply four different techniques including local swaps, minimum cuts on the boundaries of partitions, as well as contraction and dynamic programming. As our empirical study, we compare the above techniques with each other, and also to previous work in distributed graph algorithms, e.g., a label propagation method, FENNEL and Spinner. We report our results both on a private map graph and several public social networks, and show that our results beat previous distributed algorithms: e.g., compared to the label propagation algorithm, we report an improvement of 15-25% in the cut value. We also observe that our algorithms allow for scalable distributed implementation for any number of partitions. Finally, we apply our techniques for the Google Maps Driving Directions to minimize the number of multi-shard queries with the goal of saving in CPU usage. During live experiments, we observe an ~40% drop in the number of multi-shard queries when comparing our method with a standard geography-based method.

Brief Announcement: Distributed Unconstrained Local Search for Multilevel Graph Partitioning

Distributed Deep Multilevel Graph Partitioning

Local Graph Edge Partitioning with a Two-Stage Heuristic Method.

Engineering a scalable high quality graph partitioner

Enhancing Balanced Graph Edge Partition with Effective Local Search

A parallel graph partitioning algorithm to speed up the large-scale distributed graph mining.

Tera-Scale Multilevel Graph Partitioning

Two-Dimensional Balanced Partitioning and Efficient Caching for Distributed Graph Analysis

Local Graph Edge Partitioning

Scalable Edge Partitioning

Distributed Balanced Partitioning via Linear Embedding

Think Locally, Act Globally: Perfectly Balanced Graph Partitioning

Towards Fast Large-scale Graph Analysis via Two-dimensional Balanced Partitioning.

An Efficient Method for Graph Repartitioning in Distributed Environments.

A Feasible Graph Partition Framework for Parallel Computing of Big Graph

A Two-Phase Method to Balance the Result of Distributed Graph Repartitioning

A General Framework for Distributed Partitioned Optimization

A Simple Yet Effective Balanced Edge Partition Model for Parallel Computing

Bipartite-Oriented Distributed Graph Partitioning for Big Learning

Practical and high-quality partitioning algorithm for large-scale and time-evolving graphs

A Level Sensitive Cluster Based Partitioning Algorithms for Reconfigurable Systems