《ibm数据仓库与数据重整（英）.ppt》由会员分享，可在线阅读，更多相关《ibm数据仓库与数据重整（英）.ppt（52页珍藏版）》请在金锄头文库上搜索。
1、Dr. Paul F. Wilms Worldwide Data Management Sales & Technical Support,Data Warehousing & Data Re-engineering,for the BF&S Industry,DBWILMS US.IBM.COM,Why Build a Data Warehouse? Business Requirements Steps toward a Successful Implementation Data Re-engineering - Unlock the Power out of your Data Dat
2、a Transformations Data Cleansing Preparing the Data for Multi-dimensional Analysis Star Schema Representation Automating the Data Warehousing Process Scheduling of the Data Warehouse Population Monitoring the Data Warehouse Activity Metadata - The Data Warehouse Cornerstone Data to Information Proce
3、ss Metadata Models Metadata Mining Metadata Interchange Datamarts vs Enterprise-wide Data Warehouses Competitive differentiators of IBM Data Warehouse Solution,Agenda,Why Build a Data Warehouse? Business Requirements Steps toward a Successful Implementation Data Re-engineering - Unlock the Power of
4、your Data Preparing the Data for Multi-dimensional Analysis Automating the Data Warehousing Process Metadata - The Data Warehouse Cornerstone Datamarts vs Enterprise-wide Data Warehouses Competitive differentiators of IBM Data Warehouse,Agenda,Business Requirements for the BF&S Industry,Opportunity
5、questions for Data Warehouse Which customers are most likely to defect to a competitor? Who are your most profitable customers? What products do the most profitable customers use? What is the credit risk for a particular household? How risk adverse are particular segments, and why? Which additional
6、products is a customer likely to buy? Which branches should be consolidated in each geography?,Risk Management Example Need: single view of customer data detailed payment history (36 months) - see lifestyle changes acquire new customers that match good profile history data Business Benefits lower ri
7、sk exposure more cross selling opportunities accurate info leads to higher acceptance rate & more revenue Bottom Line: Loss reduction of 1.0% on a $5 billion credit base = $50M,Data Warehouse Need - Banking Industry Example,Retail Banking Transaction Data,Credit Card History Data,Customer Data,Data
8、Warehousing System,Data Warehouse,2297-6643,Credit Card Usage Need: identify customers who do NOT use their VISA card - because the bank pays $2 per card per month to VISA for the use of their logo and services How many people did not use their card in the last 12 months? Business Benefits cost redu
9、ction Bottom Line: $12M savings ( 500,000 customers x $2 x 12months),Data Warehouse Need - Banking Industry Example,7770-4505,1356-2478,4973-0037,DATA,KNOWLEDGE,DECISIONS,Patterns Trends Facts Relations Models Associations Sequences,Target Markets Funds allocation Trading options Where to advertise
10、Catalog mailing list Sales geography,Financial Economic Government Point-of-Sale Demographic Lifestyle,Why Build a Data Warehouse?,Pain: Too much Data ; we cant make the right Decision!,Subject Oriented Integrated Non-Volatile Values Over Time Supports Management Decisions,Application Oriented Limit
11、ed Integration Constantly Updated Current Values Only Supports Day-to-Day Operations,Operational Data,Informational Data,“Trust“ Accounts,“Checking“ Accounts,“Loan“ Accounts,Build a Data Warehouse to get the right Info!,Year,Month,Day,Account History,Display, Analyze, Discover,Manage and Automate,Tr
12、ansform,Distribute,Store,Extract,Find and Understand,Metadata,The Data Warehousing Process - Steps to Success,From Data to Information to Decisions,Design,Why Build a Data Warehouse? Data Re-engineering - Unlock the Power of your Data Simple Data Transformation Complex Data Transformation Data Clean
13、sing and Validation Preparing the Data for Multi-dimensional Analysis Automating the Data Warehousing Process Metadata - The Data Warehouse Cornerstone Datamarts vs Enterprise-wide Data Warehouses Competitive differentiators of IBM Data Warehouse,Agenda,Opportunity Questions for Data Transformation,
14、Is your data encoded, or explicit? Is the data valid, or missing? Is the content compatible with its description? Is there extraneous data in some fields? Is the data too detailed? Is history data available? Is there descriptive info about the data? Where? Does the data need to be assembled from var
15、ious sources? How complex are the required transformations?,Pain: we dont have the right data to run the business!,Data Transformations - An Example,Savings Account,Checking Account,Summary of Credit Card Transactions,Real Estate Loans,Personal Loans,Customer Vital Records,Consolidated Debt,Credit W
16、orthiness,Profitability,Delinquency,Risk Management,Summary of Accounts,Detailed Credit Card Transactions,Business Needs: Monitor customer credit performance Implement a customer defection prevention strategy Reduce bank reserves,Data Filtering Data Conversions Derived Columns Summaries Aggregations Table Look-up Editions,Savings Account,Chec
