Tadashi Ohmori

Fast Computation of Similarity Based on Jaccard Coefficient for Composition-Based Image Retrieval

This paper proposes an algorithm and data structure for fast computation of similarity based on Jaccard coefficient to retrieve images with regions similar to those of a query image. The similarity measures the degree of overlap between the regions of an image and those of another image. The key idea for fast computation of the similarity is to use the runlength description of an image for computing the number of overlapped pixels between the regions. We present an algorithm and data structure, and do experiments on 30,000 images to evaluate the performance of our algorithm. Experiments showed that the proposed algorithm is 5.49 (2.36) times faster than a naive algorithm on the average (the worst). And we theoretically gave fairly good estimates of the computation time.

Concurrency control of bulk access transactions on shared nothing parallel database machines

New concurrency control schemes are proposed for bulk access transactions (BATs) on shared-nothing database machines. A BAT is a transaction accessing a large bulk of data, such as a transaction updating a whole file. BATs are used for batch jobs in database services, and many BATs should be finished in a short time. Thus it is necessary to execute BATs concurrently on a database machine. When scheduling BATs, the performance is degraded by very high contention of both data and resources. Therefore, the strategy for scheduling BATs is to reduce the contention as much as possible. A weighted transaction precedence graph (WTPG) and two schedulers using it are proposed. A WTPG is used to estimate the degree of contention in a serializable schedule. Using a WTPG, the proposed schedulers reduced the contention by optimization. In the simulation, both schedulers achieve from 1.2 to 1.8 times higher throughput than atomic static lock and two-phase lock.<<ETX>>

A New Data Cube for Integrating Data Mining and OLAP

Recently there is much need of discovering useful knowledge from massive log-data which are generated by Web-based information systems. Such log-data have multiple attributes about when and where what events occurred. In order to find useful knowledge from this data-set. many aggregate queries and data-mining queries must be issued under various constraints in the when/where/what -space. To support this process. this paper proposes a new data cube named llemset Cube. This cube directly supports data-mining task in a multi-dimensional space of analysis. Definition and algorithms of the itemset cube and its application to a real data-set are described.

Scheduling batch transactions on shared-nothing parallel database machines: effects of concurrency and parallelism

Concurrency-control scheduling of batch transactions on shared-nothing (or loosely-coupled) multiprocessor database machines is discussed. Various schedulers are tested for these batch transactions to examine how well they perform when both intertransaction parallelism and intratransaction parallelism are limited. New schedulers designed for batch transaction processing are outlined which use a new tool called a weighted transaction-precedence graph (WTPG). Simulation results show that two new schedulers (globally and locally optimized WTPG schedulers) are the best performers under various workloads.<<ETX>>

An Algebraic Deductive Database Managing a Mass of Rule Clauses

This paper proposes a deductive database which manages a mass of rule-clauses in a disk as well as many fact-clauses in a disk. For this purpose, we propose a variant of relational algebra, Relational Algebra extended with Unification (RAU). Our original point consists in query-processing by RAU in this DBMS. This paper describes a compilation of a query to a RAU’s expression, optimization strategies for it, and an algorithm for a heavy RAU-operator.

Keyword search over hybrid XML-relational databases

How to realize keyword search over XML databases (XML DB) or relational databases (RDB) is a todaypsilas hot topic. In this paper, we first point out that existing keyword-search methods over those databases cannot get sufficient results. Then, we propose a new keyword-search method over hybrid XML-relational databases, and demonstrate that its answers are better than those of existing techniques. We propose a new join-operator for XML data, and utilize the new operator to enable keyword search in hybrid XML-relational databases.

Concurrency control of bulk access transactions on a parallel disk database machine: Concurrency control of bulk access transactions on a parallel disk database machine

A scheme for concurrency control of bulk access transactions on a Shared-Nothing database machine is described. A “weighted transaction precedence graph” that expresses transaction cost as well as serializability and a cautions scheduler based on it are capable of producing serializable schedules with decreased data and resource contention. In simulations the new scheduler achieved throughout levels 1.3 to 2 times those attainable with conventional locking protocols.

Concurrency control of bulk access transactions on a parallel disk database machine

A scheme for concurrency control of bulk access transactions on a Shared-Nothing database machine is described. A “weighted transaction precedence graph” that expresses transaction cost as well as serializability and a cautions scheduler based on it are capable of producing serializable schedules with decreased data and resource contention. In simulations the new scheduler achieved throughout levels 1.3 to 2 times those attainable with conventional locking protocols.