Divide and conquer approach for efficient PageRank computation

Prasanna Desikan, Nishith Pathak, Jaideep Srivastava, Vipin Kumar

Research output: Chapter in Book/Report/Conference proceedingConference contribution

3 Citations (Scopus)

Abstract

PageRank is a popular ranking metric for large graphs such as the World Wide Web. Current research techniques for improving computational efficiency of PageRank have focussed on improving the I/O cost, convergence and parallelizing the computation process. In this paper, we propose a "divide and conquer" strategy for efficient computation of PageRank. The strategy is different from contemporary improvements in that it can be combined with any existing enhancements to PageRank, giving way to an entire class of more efficient algorithms. We present a novel graph-partitioning technique for dividing the graph into subgraphs, on which computation can be performed independently. This approach has two significant benefits. Firstly, since the approach focuses on work-reduction, it can be combined with any existing enhancements to PageRank. Secondly, the proposed approach leads naturally into developing an incremental approach for computation of such ranking metrics given that these large graphs evolve over a period of time. The partitioning technique is both lossless and independent of the type (variant) of PageRank computation algorithm used. The experimental results for a static single graph (graph at a single time instance) as well as for the incremental computation in case of evolving graphs, illustrate the utility of our novel partitioning approach. The proposed approach can also be applied for the computation of any other metric based on first order Markov chain model.

Original languageEnglish
Title of host publicationICWE'06: The Sixth International Conference on Web Engineering
Pages233-240
Number of pages8
DOIs
Publication statusPublished - 2006
Externally publishedYes
EventICWE'06: 6th International Conference on Web Engineering - Palo Alto, CA
Duration: 11 Jul 200614 Jul 2006

Other

OtherICWE'06: 6th International Conference on Web Engineering
CityPalo Alto, CA
Period11/7/0614/7/06

Fingerprint

Computational efficiency
Markov processes
Costs

Keywords

  • Efficient computation
  • Graph partitioning
  • PageRank
  • Ranking measures

ASJC Scopus subject areas

  • Computer Science Applications
  • Electrical and Electronic Engineering

Cite this

Desikan, P., Pathak, N., Srivastava, J., & Kumar, V. (2006). Divide and conquer approach for efficient PageRank computation. In ICWE'06: The Sixth International Conference on Web Engineering (pp. 233-240) https://doi.org/10.1145/1145581.1145629

Divide and conquer approach for efficient PageRank computation. / Desikan, Prasanna; Pathak, Nishith; Srivastava, Jaideep; Kumar, Vipin.

ICWE'06: The Sixth International Conference on Web Engineering. 2006. p. 233-240.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Desikan, P, Pathak, N, Srivastava, J & Kumar, V 2006, Divide and conquer approach for efficient PageRank computation. in ICWE'06: The Sixth International Conference on Web Engineering. pp. 233-240, ICWE'06: 6th International Conference on Web Engineering, Palo Alto, CA, 11/7/06. https://doi.org/10.1145/1145581.1145629
Desikan P, Pathak N, Srivastava J, Kumar V. Divide and conquer approach for efficient PageRank computation. In ICWE'06: The Sixth International Conference on Web Engineering. 2006. p. 233-240 https://doi.org/10.1145/1145581.1145629
Desikan, Prasanna ; Pathak, Nishith ; Srivastava, Jaideep ; Kumar, Vipin. / Divide and conquer approach for efficient PageRank computation. ICWE'06: The Sixth International Conference on Web Engineering. 2006. pp. 233-240
@inproceedings{41bfc9f023c44a839196fc879f979ced,
title = "Divide and conquer approach for efficient PageRank computation",
abstract = "PageRank is a popular ranking metric for large graphs such as the World Wide Web. Current research techniques for improving computational efficiency of PageRank have focussed on improving the I/O cost, convergence and parallelizing the computation process. In this paper, we propose a {"}divide and conquer{"} strategy for efficient computation of PageRank. The strategy is different from contemporary improvements in that it can be combined with any existing enhancements to PageRank, giving way to an entire class of more efficient algorithms. We present a novel graph-partitioning technique for dividing the graph into subgraphs, on which computation can be performed independently. This approach has two significant benefits. Firstly, since the approach focuses on work-reduction, it can be combined with any existing enhancements to PageRank. Secondly, the proposed approach leads naturally into developing an incremental approach for computation of such ranking metrics given that these large graphs evolve over a period of time. The partitioning technique is both lossless and independent of the type (variant) of PageRank computation algorithm used. The experimental results for a static single graph (graph at a single time instance) as well as for the incremental computation in case of evolving graphs, illustrate the utility of our novel partitioning approach. The proposed approach can also be applied for the computation of any other metric based on first order Markov chain model.",
keywords = "Efficient computation, Graph partitioning, PageRank, Ranking measures",
author = "Prasanna Desikan and Nishith Pathak and Jaideep Srivastava and Vipin Kumar",
year = "2006",
doi = "10.1145/1145581.1145629",
language = "English",
isbn = "1595933522",
pages = "233--240",
booktitle = "ICWE'06: The Sixth International Conference on Web Engineering",

}

TY - GEN

T1 - Divide and conquer approach for efficient PageRank computation

AU - Desikan, Prasanna

AU - Pathak, Nishith

AU - Srivastava, Jaideep

AU - Kumar, Vipin

PY - 2006

Y1 - 2006

N2 - PageRank is a popular ranking metric for large graphs such as the World Wide Web. Current research techniques for improving computational efficiency of PageRank have focussed on improving the I/O cost, convergence and parallelizing the computation process. In this paper, we propose a "divide and conquer" strategy for efficient computation of PageRank. The strategy is different from contemporary improvements in that it can be combined with any existing enhancements to PageRank, giving way to an entire class of more efficient algorithms. We present a novel graph-partitioning technique for dividing the graph into subgraphs, on which computation can be performed independently. This approach has two significant benefits. Firstly, since the approach focuses on work-reduction, it can be combined with any existing enhancements to PageRank. Secondly, the proposed approach leads naturally into developing an incremental approach for computation of such ranking metrics given that these large graphs evolve over a period of time. The partitioning technique is both lossless and independent of the type (variant) of PageRank computation algorithm used. The experimental results for a static single graph (graph at a single time instance) as well as for the incremental computation in case of evolving graphs, illustrate the utility of our novel partitioning approach. The proposed approach can also be applied for the computation of any other metric based on first order Markov chain model.

AB - PageRank is a popular ranking metric for large graphs such as the World Wide Web. Current research techniques for improving computational efficiency of PageRank have focussed on improving the I/O cost, convergence and parallelizing the computation process. In this paper, we propose a "divide and conquer" strategy for efficient computation of PageRank. The strategy is different from contemporary improvements in that it can be combined with any existing enhancements to PageRank, giving way to an entire class of more efficient algorithms. We present a novel graph-partitioning technique for dividing the graph into subgraphs, on which computation can be performed independently. This approach has two significant benefits. Firstly, since the approach focuses on work-reduction, it can be combined with any existing enhancements to PageRank. Secondly, the proposed approach leads naturally into developing an incremental approach for computation of such ranking metrics given that these large graphs evolve over a period of time. The partitioning technique is both lossless and independent of the type (variant) of PageRank computation algorithm used. The experimental results for a static single graph (graph at a single time instance) as well as for the incremental computation in case of evolving graphs, illustrate the utility of our novel partitioning approach. The proposed approach can also be applied for the computation of any other metric based on first order Markov chain model.

KW - Efficient computation

KW - Graph partitioning

KW - PageRank

KW - Ranking measures

UR - http://www.scopus.com/inward/record.url?scp=34250625683&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=34250625683&partnerID=8YFLogxK

U2 - 10.1145/1145581.1145629

DO - 10.1145/1145581.1145629

M3 - Conference contribution

AN - SCOPUS:34250625683

SN - 1595933522

SN - 9781595933522

SP - 233

EP - 240

BT - ICWE'06: The Sixth International Conference on Web Engineering

ER -