Data warehouse methodologies

Views:
 
Category: Education
     
 

Presentation Description

No description available.

Comments

Presentation Transcript

Comparison of Vendor based Data Warehousing Methodologies : 

Comparison of Vendor based Data Warehousing Methodologies Group Members Aafia Kamal Sidra Sarwar Rana

Presentation Outline : 

Presentation Outline Abstract Introduction Literature Survey Data Warehousing Methodologies Tasks in Data Warehousing Methodology Vendors Description Attributes Description Comparison of the Vendor Attributes Conclusion

Abstract : 

Abstract A data warehouse is a subject oriented, integrated, time-variant, and nonvolatile collection of data that supports managerial decision making . Data warehousing has been cited as the highest-priority project for IT executives. This increasing demand of Data warehousing has also amplified competition in the market and there exist many vendors providing data warehousing solution.

Abstract : 

Abstract With so many vendors methodologies with different characteristics it becomes difficult for the customer to choose the ideal solution for the company . In this paper, we will review and compare several prominent vendor based data warehousing methodologies based on a common set of attributes and conclude the better vendor based solution in different scenarios.

Introduction : 

Introduction The concept of data warehousing refers to a large scale database that integrates as centralized data shared by various users thereby providing extensive analytical and predictive capabilities Data warehouses are employed in a number of industries ranging from financial institutions to telecommunications organizations. The capability of data warehouses to store data for long periods of time and ability to process huge volumes of data in a short period makes them ideal for data mining and predictive analytics.

Introduction : 

Introduction To facilitate such detailed and extensive analysis there exist different vendor based solutions, which provide different types of data warehousing methodologies and tools to support the growing market and trends. The methodologies used by these companies differ in details. Yet they all focus on the techniques of capturing and modeling user requirements in a meaningful way. In this paper we will be comparing different vendors based methodologies on the bases of different attributes.

Literature Survey : 

Literature Survey In first paper authors Analyzed different Data Warehousing Methodologies and categorize them into three types as vendor based, infrastructure based and information modeling based Data Warehousing Methodologies. Another Article they evaluated the NCR’s and IBI methodologies by comparing the two crucial phases of the data warehouse development, requirement gathering and data modeling and analysis

Literature Survey : 

Literature Survey White paper discuss the data warehouse products from three traditional vendors, as well as the newer appliance and column-based vendors. They compare the strengths and weaknesses of these products to Microsoft® SQL Server® 2008 Research papers, articles, vendor’s websites and reviews we analyzed for finding about different vendors data warehouse methodologies

Data Warehousing Methodologies : 

Data Warehousing Methodologies Core technology vendors Companies that put up for sale database engines. NCR’s Teradata , Oracle, IBM’s DB2 and Microsoft’s SQL Server based. Infrastructure vendors Provide infrastructure tools that could be a mechanism to manage metadata using repositories, to help extract, transfer, and to load data into the data warehouse, or to help create end user solutions. SAS, Informatics, Computer Associates Platinum, Visible Technologies. Information modeling vendors Focus on the techniques of capturing and modeling user requirements ERP vendors (SAP and PeopleSoft), a general business consulting company (Cap Gemini Ernst Young), and two IT/data warehouse consulting companies (Corporate Information Designs and Creative Data)

Tasks in Data Warehousing Methodology : 

Tasks in Data Warehousing Methodology Business requirements analysis Data design Architecture design Implementation strategy Deployment

Tasks in Data Warehousing Methodology : 

Tasks in Data Warehousing Methodology Business requirements analysis Techniques such as interviews, brainstorming, and JAD sessions are used to extract requirements. A conceptual model i.e. subject area data model is created for the solution of each business question which serves as the outline for the data requirements of an organization. Data design The Data design task comprises of data modeling and normalization. The two most well-liked data modeling techniques are Entity Relational modeling and Dimensional modeling.

Tasks in Data Warehousing Methodology : 

Tasks in Data Warehousing Methodology Architecture design The architecture design task includes the schema design strategy of OLTP databases. Several strategies for schema design exist, such as top down and bottom up. The data warehouse architecture design philosophies can be classified into enterprise wide data warehouse design and data mart design. Implementation strategy Inmon propose the reverse of system development life cycle. Kimball’s Business dimensional life cycle approach.

Tasks in Data Warehousing Methodology : 

Tasks in Data Warehousing Methodology Deployment The deployment task focuses on solutions for integration and data warehouse maintenance. Increased end user enhancements and repeated schema changes resulted in several versions of data warehouse.

Vendors Description : 

Vendors Description NCR/Teradata Methodology Oracle Methodology IBM DB2 Methodology Microsoft SQL Server Methodology

Attributes Description : 

Attributes Description Core Competency Requirements Modeling Data Modeling Support for Normalization/ Denormalization Architecture Design Philosophy Implementation Strategy Metadata Management Query Design Scalability Change Management

Comparison of the Vendor Attributes : 

Comparison of the Vendor Attributes Comparison of vendor based data warehousing methodologies. (NCR/ Teradata, Oracle, IBM DB2 and Microsoft SQL Server Methodology)

Conclusion : 

Conclusion We conclude with scenarios in which different vendors should be chosen from. Scenarios where organizations have large data sets to be stored, performance of the data warehouse is of essence and budget constraints are non existent. A large scale multi node Teradata system is ideally suited to such scenario. Scenarios where the dataset to be stored are large, but high performance is not mandatory, budget is constrained & extensibility of the data warehouse is not desired it would be prudent to use Oracle or IBM data warehouse.

Conclusion : 

Conclusion For medium scale data size & performance requirement IBM & Oracle can be used with normalized structure so that extensibility is ensured along side performance. In situations where the organization opting for the data warehouse has small amounts of data and has limited budget. Microsoft SQL Server is the best suited solution due to its low cost and widespread support available.

Thank you !!!!! : 

Thank you !!!!!