Database Systems Journal, Vol. IV, Issue 2/2013
|
1. ETL as a Necessity for Business Architectures (p. 3-12)Aurelian TITIRISCA, University of Economic Studies, Bucharest, Romania |
|
Today, the new trend that companies are following is to digitalize all data they have in order to reduce costs with physical space and better handle data volume. The new era, the era of bundling data sources and huge volumes of data continues this decade also. In all industries, companies are starting to understand and appreciate the value of data, the value that these volumes of data can produce.
Companies are collecting huge volumes of data but not always the actual solution for business intelligence (BI) can handle these volumes. To obtain information, those volumes of data are analyzed with extract-transform-load (ETL) software solutions. Companies, in the current economic context, are finding hard to invest in improvements of BI process including in ETL process.
In this paper I will demonstrate why this kind of investment is necessary and also I will demonstrate that ETL process must be included in BI and big data architectures. In the following pages I will refer to business architectures as BI and big data architectures.
Keywords: ETL, business intelligence, Big Data, Internet, dataset. |
2. The Development of a Benchmark Tool for NoSQL Databases (p. 13-20)Ion LUNGU, University of Economic Studies, Bucharest, RomaniaBogdan George TUDORICA, Petroleum-Gas University, Ploiesti, Romania |
|
The aim of this article is to describe a proposed benchmark methodology and software application targeted at measuring the performance of both SQL and NoSQL databases. These represent the results obtained during PhD research (being actually a part of a larger application intended for NoSQL database management). A reason for aiming at this particular subject is the complete lack of benchmarking tools for NoSQL databases, except for YCBS [1] and a benchmark tool made specifically to compare Redis to RavenDB. While there are several well-known benchmarking systems for classical relational databases (starting with the canon TPC-C, TPC-E and TPC-H), on the other side of databases world such tools are mostly missing and seriously needed.
Keywords: NoSQL database, testing, benchmark. |
3. Retrieving Targeted Results from a Static File Repository using a Keyword matching Mechanism relying on a Cluster-based Algorithm (p. 21-30)Madalina ZURINI, University of Economic Studies, Bucharest, RomaniaDiana BUTUCEA, University of Economic Studies, Bucharest, Romania |
|
Due to the fact that the web based solutions designed for learning contain, among the compulsory functions, the uploading of didactic materials of the person doing the examining (on a regular basis this being the professor) and the possibility of accessing these by the examinee (on a regular basis this being the student), within this paper we have chosen to set the goal of finding a resolution that will enable the access of content relevant to the person being examined. Hereby we have suggested a prototype which will capacitate the singling out and grouping of documents depending on the keywords, which will be followed by a visual search depending on the distance between two documents, by the recurrence of the closest k documents to the one being the element of interest - reaching the optimal alternative in case of a performance oriented point of view. The algorithm needed for the extraction of this data is presented within the paper. An optimization model is proposed in order to reduce the time consuming component in regards to the minimization of differences in the quality of the documents resulted in the automatic search using a k Nearest Neighbour grid search engine.
Keywords: e-learning, grid-based algorithm, keyword, kNN, automatic grid search optimization. |
4. Enhancing ETL Performance with Warehouse Builder (p. 31-41)Manole VELICANU, University of Economic Studies, Bucharest, RomaniaLarisa COPCEA (TEOHARI), University of Economic Studies, Bucharest, Romania |
|
We live in a dynamic world, in a permanent move, were performance of information systems continually increases. Thus, the need of being informed, regardless of place or time, is very great using data warehouses solutions. Therefore, the need for efficient information systems is very high. The efficiency can be achieved by using various methods/techniques and technologies "to build" the data warehouse. The ones, we will present in our paper, are: methods of Enhancing ETL Performance with Warehouse Builder: the purpose of ETL strategies is to create an integrated, complex, coherent software solution; techniques of data warehouse optimization: in order to improve the performance of data warehouse processing can be applied several optimization techniques; and "in terms" of technology, we will consider: data warehouses (using Oracle Warehouse Builder).
Keywords: performance, Oracle Warehouse Builder, efficiency of information systems, extraction, transformation, and loading (ETL), data warehouse optimization. |
5. A comparative Review of Extraction, Transformation and Loading Tools (p. 42-51)Amanpartap Singh PALL, School of Information Technology, APJIMTC, Jalandhar, IndiaJaiteg Singh KHAIRA, Chitkara Institute of Engineering and Technology, Rajpura, India |
|
Business today forces the enterprises to run different but coexisting information systems. However, data warehousing enterprises have a dilemma of choosing the right ETL process and the right ETL tool for their organization as one wrong step or choice may lead to a series of losses both monetarily and by time, not to mention the amount of laborious work that the workers would put in. The organization can choose from a variety of ETL tools but without exploring or the knowledge of their features this would again result in a bad decision making process. In this paper, we have tried to present a comparative review of some of the leading ETL tools just to acquaint the users with its features and drawbacks.
Keywords: ETL tools, tool comparison. |