Yon Dohn Chung

Parallel data processing with MapReduce: a survey

A prominent parallel data processing tool MapReduce is gaining significant momentum from both industry and academia as the volume of data to analyze grows rapidly. While MapReduce is used in many areas where massive data analysis is required, there are still debates on its performance, efficiency per node, and simple abstraction. This survey intends to assist the database and open source communities in understanding various technical aspects of the MapReduce framework. In this survey, we characterize the MapReduce framework and discuss its inherent pros and cons. We then introduce its optimization strategies reported in the recent literature. We also discuss the open issues and challenges raised on parallel data analysis with MapReduce.

Effective Data Placement for Wireless Broadcast

This paper investigates how to place data objects on air for wireless broadcast such that mobile clients can access the data in short latency. We first define and analyze the problem of wireless data placement, and also propose a measure, named Query Distance (QD), which represents the coherence degree of data set accessed by a query. We show that the problem is NP-complete, and then propose an effective data placement method that constructs the broadcast schedule by appending each querys data set in greedy way. We show through performance experiments that the proposed method reduces the access time of mobile query.

QEM: a scheduling method for wireless broadcast data

In mobile distributed systems the data on air can be accessed by a large number of clients. This paper describes the way clients access the wireless broadcast data with short latency. We define and analyze the problem of wireless data scheduling and we propose a measure, named QueryDistance (QD), which represents the degree of coherence for the data set accessed by a query. We give a practically usable method named QEM which constructs the broadcast schedule by expanding each querys data set in greedy way. We also evaluate the performance of our method by experiments.

SPIDER: a system for scalable, parallel / distributed evaluation of large-scale RDF data

RDF is a data model for representing labeled directed graphs, and it is used as an important building block of semantic web. Due to its flexibility and applicability, RDF has been used in applications, such as semantic web, bioinformatics, and social networks. In these applications, large-scale graph datasets are very common. However, existing techniques are not effectively managing them. In this paper, we present a scalable, efficient query processing system for RDF data, named SPIDER, based on the well-known parallel/distributed computing framework, Hadoop. SPIDER consists of two major modules (1) the graph data loader, (2) the graph query processor. The loader analyzes and dissects the RDF data and places parts of data over multiple servers. The query processor parses the user query and distributes sub queries to cluster nodes. Also, the results of sub queries from multiple servers are gathered (and refined if necessary) and delivered to the user. Both modules utilize the MapReduce framework of Hadoop. In addition, our system supports some features of SPARQL query language. This prototype will be foundation to develop real applications with large-scale RDF graph data.

An index replication scheme for wireless data broadcasting

Abstract In mobile distributed environment, data broadcasting has many applications because it has two desirable characteristics: energy efficiency and bandwidth efficiency. There have been some researches on the indexing mechanisms for wireless data broadcasting in the past. We first describe the problematic issue in the conventional index replication scheme, and then propose various index replication methods based on three criteria: accessibility , energy efficiency and adjacency . We evaluate the proposed scheme in analytic and experimental ways. We show that the proposed index replication scheme reduces the data access cost of mobile clients in an energy efficient manner. We also discuss the relaxation of the given criteria.

Energy- and Latency-Efficient Processing of Full-Text Searches on a Wireless Broadcast Stream

In wireless mobile computing environments, broadcasting is an effective and scalable technique to disseminate information to a massive number of clients, wherein the energy usage and latency are considered major concerns. This paper presents an indexing scheme for the energy- and latency-efficient processing of full-text searches over the wireless broadcast data stream. Although a lot of access methods and index structures have been proposed in the past for full-text searches, all of them are targeted for data in disk storage, not wireless broadcast channels. For full-text searches on a wireless broadcast stream, we firstly introduce a naive, inverted list-style indexing method, where inverted lists are placed in front of the data on the wireless channel. In order to reduce the latency overhead, we propose a two-level indexing method which adds another level of index structure to the basic inverted list-style index. In addition, we propose a replication strategy of the index list and index tree to further improve the latency performance. We analyze the performance of the proposed indexing scheme with respect to the latency and energy usage measures, and show the optimality of index replication. The correctness of the analysis is demonstrated through simulation experiments, and the effectiveness of the proposed scheme is shown by implementing a real wireless information delivery system.

In-Network Processing for Skyline Queries in Sensor Networks

Wireless sensor networks can be used in various fields, e.g., military and civil applications. The technique of saving energy to prolong the life of sensor nodes is one of main challenges to resource-constrained sensor networks. Therefore, in-network aggregation of data has been proposed in resource-constrained environments for energy efficiency. Most previous works on in-network aggregation only support a one-dimensional data (e.g., MIN and MAX). To support a multi-dimensional data, the skyline query is used. The skyline query returns a set of points that are not dominated by any other point on all dimensions. The majority of previous skyline query processing methods (e.g., BNL and BBS) work on centralized storage. Centralized query processing methods do not have merits in terms of energy efficiency in high event rate environments. In this paper, we propose new algorithm of in-network processing for the skyline queries. The proposed algorithm reduces the communication cost and evenly distributes load. The experimental results show the advantages of our algorithm over in-network aggregation in terms of improving energy efficiency.

Efficient stream organization for wireless broadcasting of XML data

This paper presents a wireless streaming method for XML data which supports energy-efficient processing of queries over the stream in mobile clients. We propose new stream organizations for XML data which have different kinds of addresses to related data in the stream. We describe event-driven stream generation algorithms for the proposed stream structures and provide search algorithms for simple XML path queries which leverage the access mechanisms incorporated in the stream. Experimental results show that our approach can effectively improve the tuning time performance of user queries in a wireless broadcasting environment.

Processing generalized k-nearest neighbor queries on a wireless broadcast stream

In this paper, we investigate the problem of processing generalized k-nearest neighbor (GkNN) queries, which involve both spatial and non-spatial specifications for data objects, in a wireless broadcasting system. We present a method for processing GkNN queries on the broadcast stream. In particular, we propose a novel R-tree variant index structure, called the bit-vector R-tree (bR-tree), which stores additional bit-vector information to describe non-spatial attribute values of the data objects. In addition, each node in the bR-tree stores only one pointer to its children, which makes the bR-tree compact. We generate the broadcast stream by multiplexing the bR-tree and the data objects in the broadcasting channel. The corresponding search algorithm for the broadcast stream is also described. Through a series of comprehensive simulation experiments, we prove the efficiency of the proposed method with regard to energy consumption, latency, and memory requirement, which are the major performance concerns in a wireless broadcasting system. Furthermore, we test the practicality of the proposed method in a real prototype system.

QR-tree: An efficient and scalable method for evaluation of continuous range queries

Abstract In this paper, we explore the problem of the scalable evaluation of continuous range queries (CRQs) over moving objects, each of which continually retrieves the moving objects that are currently within a given query region of interest. Most existing methods assume that moving objects continually communicate with the server to report their current locations and the server continuously updates the results of queries. However, such an assumption degrades the system performance, because the communication cost is huge and the server workload is increased when the number of moving objects and queries is enormous. In this paper, we propose a novel query indexing structure, referred to as the Query Region tree (QR-tree), which allows the server to cooperate with moving objects efficiently by leveraging the available computational resources of the moving objects to improve the overall system performance. In addition, we present another version of the QR-tree, called the Bit-vector Query Region tree (BQR-tree), for the evaluation of CRQs that specify additional non-spatial selections. The BQR-tree stores a summary of the non-spatial information specified by CRQs in the form of bit-vectors. Through a series of comprehensive simulations, we verify the efficiency of the QR-tree and the BQR-tree in terms of the communication cost and server workload.