CMU-ML-10-100
Machine Learning Department
School of Computer Science, Carnegie Mellon University



CMU-ML-10-100

Fast Nearest-neighbor Search
in Disk-resident Graphs

Purnamrita Sarkar, Andrew W. Moore

February 2010

CMU-ML-10-100.pdf


Keywords: Random walks, link prediction, external memory


Link prediction, personalized graph search, fraud detection, and many such graph mining problems revolve around the computation of the most "similar" k nodes to a given query node. One widely used class of similarity measures is based on random walks on graphs, e.g., personalized pagerank, hitting and commute times, and simrank. There are two fundamental problems associated with these measures. First, existing online algorithms typically examine the local neighborhood of the query node which can become significantly slower whenever high-degree nodes are encountered (a common phenomenon in real-world graphs). We prove that turning high degree nodes into sinks results in only a small approximation error, while greatly improving running times. The second problem is that of computing similarities at query time when the graph is too large to be memoryresident. The obvious solution is to split the graph into clusters of nodes and store each cluster on a disk page; ideally random walks will rarely cross cluster boundaries and cause page-faults. Our contributions here are twofold: (a) we present an efficient deterministic algorithm to find the k closest neighbors (in terms of personalized pagerank) of any query node in such a clustered graph, and (b) we develop a clustering algorithm (RWDISK) that uses only sequential sweeps over data files. Empirical results on several large publicly available graphs like DBLP, Citeseer and Live-Journal (∼ 90 M edges) demonstrate that turning high degree nodes into sinks not only improves running time of RWDISK by a factor of 3 but also boosts link prediction accuracy by a factor of 4 on average. We also show that RWDISK returns more desirable (high conductance and small size) clusters than the popular clustering algorithm METIS, while requiring much less memory. Finally our deterministic algorithm for computing nearest neighbors incurs far fewer page-faults (factor of 5) than actually simulating random walks.

30 pages


SCS Technical Report Collection
School of Computer Science homepage

This page maintained by reports@cs.cmu.edu