Database Systems Journal

ISSN 2069 - 3230

The journal is published under the sponsorship of
The Bucharest University of Economic Studies
and it is produced by the university's own publishing division,
The Bucharest University of Economic Studies Publishing House


Database Systems Journal, Vol. I, Issue 1/2010
Issue Topic: Database Systems


Open PDF Journal


CONTENTS


1. Spatial Operations (p. 5-8)
Anda VELICANU, Economic Informatics and Cybernetics Department, University of Economic Studies, Bucharest, Romania
This paper contains a brief description of the most important operations that can be performed on spatial data such as spatial queries, create, update, insert, delete operations, conversions, operations on the map or analysis on grid cells. Each operation has a graphical example and some of them have code examples in Oracle and PostgreSQL.
Keywords: Spatial operations, Querying operations, Spatial data.
2. Database Access Through Java Technologies (p. 9-18)
Ion LUNGU, University of Economic Studies, Bucharest, Romania
Nicolae MERCIOIU, University of Economic Studies, Bucharest, Romania
As a high level development environment, the Java technologies offer support to the development of distributed applications, independent of the platform, providing a robust set of methods to access the databases, used to create software components on the server side, as well as on the client side. Analyzing the evolution of Java tools to access data, we notice that these tools evolved from simple methods that permitted the queries, the insertion, the update and the deletion of the data to advanced implementations such as distributed transactions, cursors and batch files. The client-server architectures allows through JDBC (the Java Database Connectivity) the execution of SQL (Structured Query Language) instructions and the manipulation of the results in an independent and consistent manner. The JDBC API (Application Programming Interface) creates the level of abstractization needed to allow the call of SQL queries to any DBMS (Database Management System). In JDBC the native driver and the ODBC (Open Database Connectivity)-JDBC bridge and the classes and interfaces of the JDBC API will be described. The four steps needed to build a JDBC driven application are presented briefly, emphasizing on the way each step has to be accomplished and the expected results. In each step there are evaluations on the characteristics of the database systems and the way the JDBC programming interface adapts to each one. The data types provided by SQL2 and SQL3 standards are analyzed by comparison with the Java data types, emphasizing on the discrepancies between those and the SQL types, but also the methods that allow the conversion between different types of data through the methods of the ResultSet object. Next, starting from the metadata role and studying the Java programming interfaces that allow the query of result sets, we will describe the advanced features of the data mining with JDBC. As alternative to result sets, the Rowsets add new functionalities that enhance the flexibility of the applications. These are analyzed and the approach is described.
Keywords: Java, JDBC, Database access, SQL.
3. Optimization of Data Requests Timing by Working with Matrixes under MSAccess Environment (p. 19-22)
Alexandru ATOMEI, University of Economic Studies, Bucharest, Romania
This paper is going to emphasize an optimised code in order to manage matrix calculus under MSAccess. The economic impact of using such a method is the optimal cost-benefit solution, and optimised timing for data management. As well, matrix calculus is the base of Variance-Covariance method used by financial corporations as an advanced method for estimation of market risk movements with direct impact over the capital required by prudential bodies.
Keywords: Visual Basic, DAO (Data Access Objects) Recordset, System DSN (Data Source Name) driver, Variance-Covariance Matrix, Value at Risk.
4. SEO Techniques for Business Websites (p. 23-26)
Alexandru ENACEANU, Romanian-American University, Bucharest, Romania
In the world of website marketing, search engines are an essential key to success. They are the most important way to bring traffic to websites. Understanding how search engines work and what they require is an important first step to harnessing their marketing power. There are proven methods to search engine marketing involving website design, content adaptation, and keyword strategy. The primary goal of these methods is to bring traffic to your site. The secondary goal is for that traffic to be targeted to your product. In the internet marketing game, exposure is essential. But marketing efficiency requires effective exposure to the right prospects.
Keywords: SEO, search engine optimization, pagerank, business website, Internet.
5. Solutions for improving data extraction from virtual data warehouses (p. 27-36)
Adela BARA, University of Economic Studies, Bucharest, Romania
The data warehousing project’s team is always confronted with low performance in data extraction. In a Business Intelligence environment this problem can be critical because the data displayed are no longer available for taking decisions, so the project can be compromised. In this case there are several techniques that can be applied to reduce queries’ execution time and to improve the performance of the BI analyses and reports. Some of the techniques that can be applied to reduce the cost of execution for improving query performance in BI systems will be presented in this paper.
Keywords: Virtual data warehouse, Data extraction, SQL tuning, Query performance.
6. The Optimization of Algorithms in the Process of Temporal Data Mining Using the Compute Unified Device Architecture (p. 37-47)
Alexandru PIRJAN, University of Economic Studies, Bucharest, Romania
Considering the importance and usefulness of real time data mining, in recent years the concern of researchers to discover new hardware architectures that can manage and process large volumes of data has increased significantly. In this paper the performance of algorithms for temporal data mining that are implemented in the new Compute Unified Device Architecture (CUDA) from the latest generation of graphics processing units (GPU) will be analyzed and reviewed. The performance will be evaluated taking into account the type of algorithm, data access, the problems' size, the GPU’s processor generation, the number of threads processed.
Keywords: Temporal data mining, MapReduce, CUDA, GPU, Fermi, thread, kernel.