Parallel Data Mining Research Articles

BPGM: A big graph mining tool

The design and implementation of a scalable parallel mining system target for big graph analysis has proven to be challenging. In this study, we propose a parallel data mining system for analyzing big graph data generated on a Bulk Synchronous Parallel (BSP) computing model named BSP-based Parallel Graph Mining (BPGM). This system has four sets of parallel graph mining algorithms programmed in the BSP parallel model and a well-designed workflow engine optimized for cloud computing to invoke these algorithms. Experimental results show that the graph mining algorithm components in BPGM are efficient and have better performance than big cloud-based parallel data miner and BC-BSP.

Scolopax

The formulation of hypotheses based on patterns found in data is an essential component of scientific discovery. As larger and richer data sets become available, new scalable and user-friendly tools for scientific discovery through data analysis are needed. We demonstrate Scolopax, which explores the idea of a search engine for hypotheses. It has an intuitive user interface that supports sophisticated queries. Scolopax can explore a huge space of possible hypotheses, returning a ranked list of those that best match the user preferences. To scale to large and complex data sets, Scolopax relies on parallel data management and mining techniques. These include model training, efficient model summary generation, and novel parallel join techniques that together with traditional approaches such as clustering manipulate massive model-summary collections to find the most interesting hypotheses. This demonstration of Scolopax uses a real observational data set, provided by the Cornell Lab of Ornithology. It contains more than 3.3 million bird sightings reported by citizen scientists and has almost 2500 attributes. Conference attendees have the opportunity to make novel discoveries in this data set, ranging from identifying variables that strongly affect bird populations in specific regions to detecting more sophisticated patterns such as habitat competition and migration.

Dynamic group communication for large-scale parallel data mining

Exascale systems are the next frontier in high-performance computing and are expected to deliver a performance of the order of 1018 operations per second using massive multicore processors. Very large- and extreme-scale parallel systems pose critical algorithmic challenges, especially related to concurrency, locality and the need to avoid global communication patterns. This work investigates a novel protocol for dynamic group communication that can be used to remove the global communication requirement and to reduce the communication cost in parallel formulations of iterative data mining algorithms. The protocol is used to provide a communication-efficient parallel formulation of the k-means algorithm for cluster analysis. The approach is based on a collective communication operation for dynamic groups of processes and exploits non-uniform data distributions. Non-uniform data distributions can be either found in real-world distributed applications or induced by means of multidimensional binary search trees. The analysis of the proposed dynamic group communication protocol has shown that it does not introduce significant communication overhead. The parallel clustering algorithm has also been extended to accommodate an approximation error, which allows a further reduction of the communication costs. The effectiveness of the exact and approximate methods has been tested in a parallel computing system with 64 processors and in simulations with 1024 processing elements.

Non-uniform data distribution for communication-efficient parallel clustering

Global communication requirements and load imbalance of some parallel data mining algorithms are the major obstacles to exploit the computational power of large-scale systems. This work investigates how non-uniform data distributions can be exploited to remove the global communication requirement and to reduce the communication cost in parallel data mining algorithms and, in particular, in the k-means algorithm for cluster analysis. In the straightforward parallel formulation of the k-means algorithm, data and computation loads are uniformly distributed over the processing nodes. This approach has excellent load balancing characteristics that may suggest it could scale up to large and extreme-scale parallel computing systems. However, at each iteration step the algorithm requires a global reduction operation which hinders the scalability of the approach. This work studies a different parallel formulation of the algorithm where the requirement of global communication is removed, while maintaining the same deterministic nature of the centralised algorithm. The proposed approach exploits a non-uniform data distribution which can be either found in real-world distributed applications or can be induced by means of multi-dimensional binary search trees. The approach can also be extended to accommodate an approximation error which allows a further reduction of the communication costs. The effectiveness of the exact and approximate methods has been tested in a parallel computing system with 64 processors and in simulations with 1024 processing elements.

Research on Parallel Classification Algorithms for Large-scale Data

Because of the growing mass of data and the requirements of data mining's individuation, the traditional centralized data mining method can't adapt to this kind of demand. Cloud computing provided a cheap solution for massive data storage, analysis and handling. In order to achieve the purpose of parallel data mining in cloud environment, an improved algorithm based on the traditional Naive Bayes has been proposed in this paper. First, proposing the designing ideas of the improved algorithm in MapReduce programming model. Then using the actual data to test the algorithm. The experimental result validated that the new algorithm has higher performance and better scalability.

Distributed data mining: a survey

Most data mining approaches assume that the data can be provided from a single source. If data was produced from many physically distributed locations like Wal-Mart, these methods require a data center which gathers data from distributed locations. Sometimes, transmitting large amounts of data to a data center is expensive and even impractical. Therefore, distributed and parallel data mining algorithms were developed to solve this problem. In this paper, we survey the-state-of-the-art algorithms and applications in distributed data mining and discuss the future research opportunities.

Improving the scalability of ILP-based multi-relational concept discovery system through parallelization

Due to the increase in the amount of relational data that is being collected and the limitations of propositional problem definition in relational domains, multi-relational data mining has arisen to be able to extract patterns from relational data. In order to cope with intractably large search space and still to be able to generate high-quality patterns, ILP-based multi-relational data mining and concept discovery systems employ several search strategies and pattern limitations. Another direction to cope with the large search space is using parallelization. By parallel data mining, improvement in time efficiency and scalability can be provided without further limiting the language patterns. In this work, we describe a method for concept discovery with parallelization on an ILP-based concept discovery system. The non-parallel algorithm, namely Concept Rule Induction System (CRIS), is modified in such a way that the parts that involve high amount of query processing, which causes bottleneck, are reorganized in a data parallel way. The resulting algorithm is called, Parallel CRIS (pCRIS). A set of experiments is conducted in order to evaluate the performance of the proposed method.

Parallel data mining techniques on Graphics Processing Unit with Compute Unified Device Architecture (CUDA)

Recent development in Graphics Processing Units (GPUs) has enabled inexpensive high performance computing for general-purpose applications. Compute Unified Device Architecture (CUDA) programming model provides the programmers adequate C language like APIs to better exploit the parallel power of the GPU. Data mining is widely used and has significant applications in various domains. However, current data mining toolkits cannot meet the requirement of applications with large-scale databases in terms of speed. In this paper, we propose three techniques to speedup fundamental problems in data mining algorithms on the CUDA platform: scalable thread scheduling scheme for irregular pattern, parallel distributed top-k scheme, and parallel high dimension reduction scheme. They play a key role in our CUDA-based implementation of three representative data mining algorithms, CU-Apriori, CU-KNN, and CU-K-means. These parallel implementations outperform the other state-of-the-art implementations significantly on a HP xw8600 workstation with a Tesla C1060 GPU and a Core-quad Intel Xeon CPU. Our results have shown that GPU + CUDA parallel architecture is feasible and promising for data mining applications.

SeqGene: a comprehensive software solution for mining exome- and transcriptome- sequencing data

BackgroundThe popularity of massively parallel exome and transcriptome sequencing projects demands new data mining tools with a comprehensive set of features to support a wide range of analysis tasks.ResultsSeqGene, a new data mining tool, supports mutation detection and annotation, dbSNP and 1000 Genome data integration, RNA-Seq expression quantification, mutation and coverage visualization, allele specific expression (ASE), differentially expressed genes (DEGs) identification, copy number variation (CNV) analysis, and gene expression quantitative trait loci (eQTLs) detection. We also developed novel methods for testing the association between SNP and expression and identifying genotype-controlled DEGs. We showed that the results generated from SeqGene compares favourably to other existing methods in our case studies.ConclusionSeqGene is designed as a general-purpose software package. It supports both paired-end reads and single reads generated on most sequencing platforms; it runs on all major types of computers; it supports arbitrary genome assemblies for arbitrary organisms; and it scales well to support both large and small scale sequencing projects. The software homepage is http://seqgene.sourceforge.net.

Scaling Data Mining Algorithms to Large and Distributed Datasets

In the contemporary world of global economy real-life data is distributed and evolving consistently. For the purpose of data mining, the large set of evolving and distributed data can be handled efficiently by Parallel Data mining and Distributed Data Mining, Incremental Data mining. In this paper, we discuss about the issues and the present research work that is being carried out on parallel and distributed data mining. Adaptability of some core data mining algorithms such as decision trees, discovery of frequent patterns, clustering, etc., for parallel processing and contemporary research work related to parallel processing of the algorithms is also discussed. We have identified two approaches for carrying out distributed data mining and tried to bring out the advantages of using mobile agents in client server-based approaches, in terms of bandwidth usage and network latency.

A Study of Parallel Data Mining in a Peer-to-Peer Network

In this article, a parallel data mining algorithm in a distributed Peer-to-Peer (P2P) network is designed and proposed. The algorithm has the following advanced features: the implementation of the algorithm for all nodes in a P2P network is the same which satisfies not only the distribution but also the random walking in/out features of a P2P network; it balances the working load of each node in the P2P network well; it is easy for the maintenance and reuse of the codes. All processes of this algorithm are executed in parallel over a P2P network to reach high efficiency, fine scalability and efficient communication. Data mining for large and distributed databases in P2P networks requires more efficient parallel or distributed algorithms. Dealing with a fast changing P2P environment also demands more flexible and scaleable methods. Our parallel algorithm provides a good solution. Parallel P2P data mining applications may play a key role in the next generation of distributed database networks, file sharing networks, and search engines.

Promoting performance and separation of concerns for data mining applications on the grid

Grid Computing brought the promise of making high-performance computing cheaper and more easily available than traditional supercomputing platforms. Such a promise was very well received by the data mining (DM) community, as DM applications typically process very large datasets and are thus very resource intensive. However, since the Grid is very dynamic and parallel data mining is prone to load unbalancing, obtaining good data mining performance on the Grid is hard. It typically requires the scheduler to understand the inner workings of the application, bringing two related problems. First, good Grid schedulers tend to be very specialized in the application they target. Second, changing the application may require changing the scheduler, which may be especially challenging when there is no clear separation between the application and the scheduler code. We here propose and evaluate a knowledge-based approach that provides abstractions to the DM developer and optimizes at runtime the DM application on the Grid.

TeraScope: distributed visual data mining of terascale data sets over photonic networks

TeraScope is a framework and a suite of tools for interactively browsing and visualizing large terascale data sets. Unique to TeraScope is its utilization of the Optiputer paradigm to treat distributed computer clusters as a single giant computer, where the dedicated optical networks that connect the clusters serve as the computer’s system bus. TeraScope explores one aspect of the Optiputer architecture by employing a distributed pool of memory, called LambdaRAM, that serves as a massive data cache for supporting parallel data mining and visualization algorithms.

Effect of data skewness and workload balance in parallel data mining

To mine association rules efficiently, we have developed a new parallel mining algorithm FPM on a distributed share-nothing parallel system in which data are partitioned across the processors. FPM is an enhancement of the FDM algorithm, which we previously proposed for distributed mining of association rules (Cheung et al., 1996). FPM requires fewer rounds of message exchanges than FDM and, hence, has a better response time in a parallel environment. The algorithm has been experimentally found to outperform CD, a representative parallel algorithm for the same goal (Agrawal and Srikant, 1994). The efficiency of FPM is attributed to the incorporation of two powerful candidate sets pruning techniques: distributed and global prunings. The two techniques are sensitive to two data distribution characteristics, data skewness, and workload balance. Metrics based on entropy are proposed for these two characteristics. The prunings are very effective when both the skewness and balance are high. In order to increase the efficiency of FPM, we have developed methods to partition a database so that the resulting partitions have high balance and skewness. Experiments have shown empirically that our partitioning algorithms can achieve these aims very well, in particular, the results are consistently better than a random partitioning. Moreover, the partitioning algorithms incur little overhead. So, using our partitioning algorithms and FPM together, we can mine association rules from a database efficiently.

Systems support for scalable data mining

William A. Maniatty Department of Computer Science University at Albany Albany, NY 12222 maniatty@cs.albany.edu Mohammed J. Zaki Department of Computer Science Rensselaer Polytechnic Institute Troy, NY 12180 zaki@cs.rpi.edu ABSTRACT The urrent generation of data mining tools have limited apa ity and performan e, sin e these tools tend to be sequential. This paper explores a migration path out of this bottlene k by onsidering an integrated hardware and software approa h to parallelize data mining. Our analysis shows that parallel data mining solutions require the following omponents: parallel data mining algorithms, parallel and distributed data bases, parallel le systems, parallel I/O, tertiary storage, management of online data, support for heterogeneous data representations, se urity, quality of servi e and pri ing metri s. State of the art te hnology in these areas is surveyed with an eye towards an integration strategy leading to a omplete solution. General Terms S alable Knowledge Dis overy and Data Mining

Optimizing transport protocol parameters for large scale PC cluster and its evaluation with parallel data mining

Recently, PC clusters have come to be studied intensively for large scale parallel computers of the next generation. ATM technology is a strong candidate as a de facto standard of high speed communication networks. Therefore, an ATM-connected PC cluster is a promising platform from the cost/performance point of view, as a future high performance computing environment. Data intensive applications, such as data mining and ad hoc query processing in databases, are considered very important for massively parallel processors, as well as for conventional scientific calculations. Thus, investigating the feasibility of applications on an ATM-connected PC cluster is meaningful. In this paper, an ATM-connected PC cluster consisting of 100 PCs is reported, and characteristics of a transport layer protocol for the PC cluster are evaluated. Point-to-point communication performance is measured and discussed, when a TCP window size parameter is changed. Parallel data mining is implemented and evaluated on the cluster. Retransmission caused by cell loss at the ATM switch is analyzed, and parameters of retransmission mechanism suitable for parallel processing on the large scale PC cluster are clarified. Default TCP protocol cannot provide good performance, since a lot of collisions happen during all-to-all multicasting executed on the large scale PC cluster. Using TCP parameters with the proposed optimization, performance improvement is achieved for parallel data mining on 100 PCs.

Scalable parallel data mining for association rules

The authors propose two new parallel formulations of the Apriori algorithm (R. Agrawal and R. Srikant, 1994) that is used for computing association rules. These new formulations, IDD and HD, address the shortcomings of two previously proposed parallel formulations CD and DD. Unlike the CD algorithm, the IDD algorithm partitions the candidate set intelligently among processors to efficiently parallelize the step of building the hash tree. The IDD algorithm also eliminates the redundant work inherent in DD, and requires substantially smaller communication overhead than DD. But IDD suffers from the added cost due to communication of transactions among processors. HD is a hybrid algorithm that combines the advantages of CD and DD. Experimental results on a 128-processor Cray T3E show that HD scales just as well as the CD algorithm with respect to the number of transactions, and scales as well as IDD with respect to increasing candidate set size.

Advances and Research Directions in Data-Warehousing Technology

Information is one of the most valuable assets of an organisation and when used properly can assist in intelligent decision making that can significantly improve the functioning of an organisation. Data Warehousing is a recent technology that allows information to be easily and efficiently accessed for decision-making activities by collecting data from many operational, legacy and possibly heterogeneous data sources. On-Line Analytical Processing (OLAP) tools are well-suited for complex data analysis, such as multi-dimensional data analysis, and to assist in decision support activities while data mining tools take the process one step further and actively search the data for patterns and hidden knowledge in the data held in the warehouse. Many organisations are building, or are planning to develop, a data warehouse for their operational and decision support needs. In this paper, we present an overview of data warehousing, multi-dimensional databases, OLAP and data mining technology and discuss the directions of current research in the area. We also discuss recent developments in data warehouse modelling, view selection and maintenance, indexing schemes, parallel query processing and data mining issues. A number of technical issues for exploratory research are presented and possible solutions are also discussed.

Implementation of parallel data mining on an ATM-connected PC cluster and performance analysis of TCP retransmission mechanisms

A recent tendency in parallel computer design has been to use general-purpose components for system configuration elements such as CPUs, disks, and memories, which used to be specially developed. Although the connection network between the processors has been specially developed, it is now possible to configure a large-scale PC cluster with good performance at low cost by making use of an ATM network as a processor connection network because of the development and cost reduction of ATM network technologies in the communication field. In this paper, a large-scale PC cluster is constructed by connecting 100 personal computers by means of a general-purpose ATM network. Applications to parallel data mining are evaluated and discussed. In particular, an analysis is carried out with a focus on the effect of TCP retransmission with cell discarding of the ATM switch on the performance. The parameter setting of a retransmission mechanism suitable for the parallel processing in the cluster is found. Further, by developing a method for setting the retransmission spacing parameters to random values for each node, it is shown that a further improvement is possible. © 1999 Scripta Technica, Electron Comm Jpn Pt 1, 82(7): 12–24, 1999

Implementation of parallel data mining on an ATM‐connected PC cluster and performance analysis of TCP retransmission mechanisms

Implementation of parallel data mining on an ATM‐connected PC cluster and performance analysis of TCP retransmission mechanisms