Approaches to Parallel Graph-Based Knowledge Discovery

doi:10.1006/jpdc.2000.1696

Journal of Parallel and Distributed Computing

Volume 61, Issue 3, March 2001, Pages 427-446

https://doi.org/10.1006/jpdc.2000.1696 Get rights and content

Abstract

The large amount of data collected today is quickly overwhelming researchers' abilities to interpret the data and discover interesting patterns. Knowledge discovery and data mining systems contain the potential to automate the interpretation process, but these approaches frequently utilize computationally expensive algorithms. In particular, scientific discovery systems focus on the utilization of richer data representation, sometimes without regard for scalability. This research investigates approaches for scaling a particular knowledge discovery–data mining system, S $ubdue$ , using parallel and distributed resources. S $ubdue$ has been used to discover interesting and repetitive concepts in graph-based databases from a variety of domains, but requires a substantial amount of processing time. Experiments that demonstrate scalability of parallel versions of the S $ubdue$ system are performed using CAD circuit databases, satellite images, and artificially-generated databases, and potential achievements and obstacles are discussed.

References (30)

G. Karypis et al.
Multilevel k-way partitioning scheme for irregular graphs
J. Parallel Distrib. Comput
(1998)
L. Asker et al.
Ensembles as a sequence of classifiers
Proceedings of the Fifteenth International Joint Conference on Artificial Intelligence
(1997)
P. Chan et al.
Toward parallel and distributed learning by meta-learning
Working Notes of the AAAAI-93 Workshop on Knowledge Discovery in Databases
(1993)
P. Cheeseman et al.
Bayesian classification (AutoClass): Theory and results
S. Clearwater et al.
A tool for knowledge-based introduction
Proceedings of the Second International IEEE Conference on Tools for Artificial Intelligence
(1990)
D. Conklin et al.
Spatial analogy and subsumption
Proceedings of the Machine Learning Conference
(1992)
D.J. Cook et al.
Substructure discovery using minimum description length and background knowledge
J. Artificial Intelligent Res
(1994)
D. J. Cook, and, L. B. Holder, Graph-based data mining, IEEE Intelligent Systems, in...
D.J. Cook et al.
Scalable discovery of informative structural concepts using domain knowledge
IEEE Expert
(1996)
D.J. Cook et al.
Maximizing the benefits of parallel search using machine learning
Proceedings of the National Conference on Artificial Intelligence
(1997)

D.J. Cook et al.

Adaptable incremental deepening search

J. Artificial Intelligence Research

(1998)

S. Djoko

The Role of Domain Knowledge in Substructure Discovery

(Aug. 1995)

S. Djoko et al.

Analyzing the benefits of domain knowledge in substructure discovery

Proceedings of the First International Conference on Knowledge Discovery and Data Mining

(1995)

U.M. Fayyad et al.

From data mining to knowledge discovery: An overview

Cited by (38)

High performance GPU primitives for graph-tensor learning operations
2021, Journal of Parallel and Distributed Computing
Citation Excerpt :
The entire cuGraph-Tensor library is running on top of NVIDIA CUDA [7] libraries including cuBLAS and cuSolver, and existing libraries including Magma [1] and KBLAS [4] to support upper applications. These graph-tensor operations are widely used in various applications such as video compression [35], data completion [16], data compression [30], machine learning [19] and pattern recognition [12], feature learning [2], knowledge discovery [5] and network analysis [30]. We develop a graph data completion application to demonstrate the usage of the cuGraph-Tensor library in Section 4.3.
Graph-tensor learning operations extend tensor operations by taking the graph structure into account, which have been applied to diverse domains such as image processing and machine learning. However, the running time of graph-tensor operations increases rapidly with the number of nodes and the dimension of data on nodes, making them impractical for real-time applications. In this paper, we propose a GPU library called cuGraph-Tensor for high-performance graph-tensor learning operations, which consists of eight key operations: graph shift (g-shift), graph Fourier transform (g-FT), inverse graph Fourier transform (inverse g-FT), graph filter (g-filter), graph convolution (g-convolution), graph-tensor product (g-product), graph-tensor SVD (g-SVD) and graph-tensor QR (g-QR). cuGraph-Tensor supports scalar, vector, and matrix data processing on each graph node. We propose optimization techniques on computing, memory accesses, and CPU–GPU communications that significantly improve the performance of the graph-tensor learning operations. Using the optimized operations, cuGraph-Tensor builds a graph data completion application for fast and accurate reconstruction of incomplete graph data. In the experiments, the proposed graph learning operations achieve up to $142.12 \times$ speedups versus CPU-based GSPBOX and CPU MATLAB implementations running on two Xeon CPUs. The graph data completion application achieves up to $174.38 \times$ speedups over the CPU MATLAB implementation, and up to $3.82 \times$ speedups with better accuracy over the GPU-based tensor completion in the cuTensor-tubal library.
Secure Outsourcing for Normalized Cuts of Large-Scale Dense Graph in Internet of Things
2022, IEEE Internet of Things Journal
Parallelizing Automatic Temporal Cognitive Tool for Large-Scale Online Learning Analytics
2021, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Motif discovery algorithms in static and temporal networks: A survey
2020, Journal of Complex Networks
An Efficient Distributed Algorithm for Large-Scale Graph Data Mining Based on Decoupled Summary Subgraph
2020, Jisuanji Xuebao/Chinese Journal of Computers
Motif Discovery Algorithms in Static and Temporal Networks: A Survey
2020, arXiv

View all citing articles on Scopus

^☆: This work is supported by NSF Grant IRI-9502260.

^f1: E-mail: [email protected], [email protected], [email protected], [email protected]

View full text

Regular ArticleApproaches to Parallel Graph-Based Knowledge Discovery☆

Abstract

J. Parallel Distrib. Comput

Ensembles as a sequence of classifiers

Proceedings of the Fifteenth International Joint Conference on Artificial Intelligence

Toward parallel and distributed learning by meta-learning

Working Notes of the AAAAI-93 Workshop on Knowledge Discovery in Databases

Bayesian classification (AutoClass): Theory and results

A tool for knowledge-based introduction

Proceedings of the Second International IEEE Conference on Tools for Artificial Intelligence

Spatial analogy and subsumption

Proceedings of the Machine Learning Conference

Substructure discovery using minimum description length and background knowledge

J. Artificial Intelligent Res

Scalable discovery of informative structural concepts using domain knowledge

IEEE Expert

Maximizing the benefits of parallel search using machine learning

Proceedings of the National Conference on Artificial Intelligence