Question

Q 1.Describe points to be considered for implementing a web-enableddata warehouse.

Q 2.Using the knowledge you have learned in this subject, describe thefollowing

approaches to develop a datawarehouse:

•The top downapproach

•The Bottom-UpApproach

•Federatedapproach

HybridApproach

Answer 1

2)
Data warehouse is a repository of an organization'sstored data. To develop a data warehouse thenfollow below approaches:Top-Down approach:

-->The Top-down approach to data warehouse design, in which thedata warehouse is designed using a normalized enterprise datamodel.
-->In the ‘Inmon vision’ the data warehouse is atthe center of the "Corporate Information Factory" (CIF), whichprovides a logical framework fordelivering business intelligence(BI) and business management capabilities.
-->‘Inmon states’ that the data warehouse is followsbelow steps:Subject-oriented:The data in the data warehouse is organized so that all the dataelements relating tothe same real-world event or object are linkedtogether.Time-variant:The changes to the data in the data warehouse are tracked andrecorded so that reportscan be produced showing changes overtime.Non-volatile:Data inthe data warehouse is never over-written or deleted - oncecommitted, the data is static,read-only, and retained for futurereporting.Integrated:The data warehouse contains data from most or all of anorganization'soperational systems and this data is madeconsistent.
-->The top-down approach advantage is Separating the low levelwork from the higherlevel objects leads to a modular design.Modular design meansdevelopment can be self contained.-->The top-down approach often uses the traditionalworkshop or micro fabrication methods where externally-controlledtools are used to cut, mill, andshape materials into the desiredshape and order and easy to maintain.
Bottom-up approach:
-->The Bottom-up approach “data marts” are firstcreated to provide reporting and analytical capabilities forspecific business processes.
-->Data marts contain atomic data, summarized data. These datamarts can eventually be unioned together to create a comprehensivedatawarehouse.
-->The ‘data warehouse bus architecture’ isfundamental to maintaining the integrity of the data warehouse.
-->‘The bus structure’ contained all the commonelements that are used by data marts such as conformed dimensions,measures.-->This architecture makes the data warehouse more of avirtual reality than a physical reality.-->All data marts could be located in one server or couldbe located on different servers across the enterprise while thedata warehouse would be avirtual entity being nothing more than asum total of all the data marts.-->The Bottom-up approach is Faster and easierimplementation, less risk of failure and inherentlyincremental.Bottom-up approaches often use the chemical properties ofsingle molecules to cause single-molecule components toself-organize or self-assembleinto some useful conformation,(or) rely on positional assembly.
Federated approach:-->This is a hub-and –spoke architecture oftendescribed as the “architecture of architectures”.Itrecommends an integrationdatawarehouses,datamartsandpackagedapplicationsthatalreadyexistintheenterprise.-->Thegoalistointegrateexistinganalyticstructureswhereverpossibleandtodefinethe“highestvalue”metrics,dimensionsand
measuresandshareandreusethemwithinexistinganalyticstructures.-->Thismayresultinthecreationofacommonstagingareatoeliminateredundantdatafeedsorbuildingofadatawarehouse
thatsourcesdatafrommultipledatamarts,datawarehousesoranalyticapplications.
Hybridapproach:
-->The Hybridapproach aims to harness the speed and user orientation oftheBottom up approach to the integration of the top-downapproach.-->The Hybrid approach begins with an Entity Relationshipdiagram of the data marts and a gradual extension of the data martsto extend the enterprisemodel in a consistent, linearfashion.-->These data marts are developed using the star schema ordimensional models.-->The Extract, Transform and Load (ETL) tool is deployedto extract data from the source into a non persistentstaging area and then intodimensional data marts thatcontain both atomic and summary data.-->The data from the various data marts are thentransferred to the data warehouse and query tools are reprogrammedto request summary data from themarts and atomic data from thedata warehouse.-->The major benefit of a hybrid approach is that itcombines rapid development techniques within an enterprisearchitecture framework. It develops anenterprise data modeliteratively and only develops a heavyweight infrastructure onceit’s really needed.Its helpful to you.........