Building the Data Warehouse

Скачать в pdf «Building the Data Warehouse»


standard requirements development for reports



for each analysis



IND1


determine data needed



program to extract data



IND3



program to merge, analyze, combine with other data



analyze data


1ND5


1ND6


answer


question


institutionalize?



Departmental


for each subject




APP




Operational Data


Data Warehouse


Departmental Data


Individual Data


•    built application by application


•    current value data


•    data can be updated


•    online, transaction oriented


•    integrated data


•    a perspective over time


•    no update, load only


•    no online access, batch only


•    subject oriented


•    different levels of granularity


•    can contain external data


•    mostly primitive, with some public derived data


•    nonredundant


•    as detailed data over time


•    sometimes called “atomic” data


•    repetitive


•    parochial


•    summarizations, subsets


•    mixture of primitive and derived


•    parochially managed databases


•    trend analysis, demographic analysis, exception reporting, monthly key figures, etc.


•    nonrepetitive


•    temporary


•    ad hoc


•    PC oriented


•    heuristic


•    analytical


•    mostly derived data


•    limited amounts of data


•    no update of data


Figure A.4 METH 4. Data-Driven development methodology.


APPENDIX




interviews data gathering



strategic plan existing systems



technical


environment


established



for each subject

Operational Data


account

Скачать в pdf «Building the Data Warehouse»