Transcription
Data Warehouse ModelingIndustry ModelsModeling Techniques come from Mars andIndustry Models come from Venus?Maarten Ketelaars
AgendaIntroductionHigh level architectureTechnical AspectsFunctional AspectsOverview of Industry Models (Including Mapping on High level architecture)IBMTeradataSASOracleTechnical aspects of Industry ModelsBest Practices
AgendaIntroductionHigh level architectureTechnical AspectsFunctional AspectsOverview of Industry Models (Including Mapping on High level architecture)IBMTeradataSASOracleTechnical aspects of Industry ModelsBest Practices
EmployersMaarten KetelaarsManaging ConsultantNippur (2012 ---)06-18537914Maarten.Ketelaars@nippur.nlManaging ConsultantAccenture Technology Solutions (2007-2012 )BI ManagerProfessional Background and Skills Business Intelligence Professional at Nippur, responsible for allbusiness in the region Arnhem / Apeldoorn.DNV – CIBIT (2005-2007) Experienced in Data Integration, Data Migration and BusinessIntelligence.Senior Advisor / Trainer Focuses on consulting and implementation projects.SNS Bank (2001-2005) Training and coaching experience.Senior Data Architect Holds a Master of Computer Science from University ofEindhoven. Skill Set Data modeling: Data Vault, Dimensional Model,Industry models (IIW & FS-LDM, DDS) Data architecture BI Project ManagementCMG (1996-2001)BI ConsultantEconomic Institute Tilburg (1993–1996)SAS Developer4
Maarten KetelaarsManaging S - DDS(At Insurance company)EmployersNippur (2012 ---)Implemented as specifiedManaging ConsultantAccenture Technology Solutions (2007-2012 )BI ManagerProfessional Background and Skills Business Intelligence Professional at Nippur, responsible for allbusiness in the region Arnhem / Apeldoorn.Oracle –OBIEE AppsSenior Advisor / Trainer (At Telco company)DNV – CIBIT (2005-2007)IBM – IIW Experienced in Data Integration, Data Migration and Business(At Insurance company)Intelligence. Focuses on consulting and implementation projects.SNS Bank (2001-2005) Training and coaching experience.Senior Data ArchitectImplemented as specified Holds a Master of Computer Science from University ofEindhoven. Skill Set Data modeling: Data Vault, Dimensional Model,Teradata – FS/LDMIndustry models (IIW & FS-LDM, DDS) Data architectureUsedreferencemodel, BIasProjectManagementInvestigated for feasibilityCMG (1996-2001)BI ConsultantIBM - FSDM / BDWMEconomic Institute Tilburg (1993–1996)Used as reference model,especially for RT IntegrationImplemented with Data VaulttechniqueSAS Developer5
AgendaIntroductionHigh level architectureTechnical AspectsFunctional AspectsOverview of Industry Models (Including Mapping on High level architecture)IBMTeradataSASOracleTechnical aspects of Industry ModelsBest Practices
High level architectureOverview of main functional componentsSource SystemsReportingDatamartAnalyticsData WareHouseExternal
High level architectureTechnical perspectiveModeling techniques focus on HOW the Data Warehouse & Data Mart are modeledTechnical aspects of DataWarehouse modelsHistory handlingSurrogate key generationLinking between entities Attention for automatic generation of modelsFocus on automation of (ETL-)processesOften initiated by Data Modeling & DWH specialists
High level architectureFunctional perspectiveIndustry Models focus on WHAT content must be captured by the Data Warehouse & DataMartFunctional aspects of DataWarehouse modelsWhich entities must be in the data warehouseSpecific functional areas coveredUsage of the information Focus on definition of business termspre-defined entitiesOften initiated by Business architects
AgendaIntroductionHigh level architectureTechnical AspectsFunctional AspectsOverview of Industry Models (Including Mapping on High level architecture)IBMTeradataSASOracleTechnical aspects of Industry ModelsBest Practices
Overview of Industry ModelsIBM Banking Data WarehouseIBM Banking Process and Service ModelsIBM Insurance Information WarehouseIBM Insurance Process and Service ModelsIBM Health Plan Data ModelIBM Retail Data WarehouseIBM Telecommunications Data WarehouseIBM Financial Markets Data WarehouseIBM Banking and Financial Markets Data Warehouse
High level architectureMapping IIW on HLASource SystemsReportingDatamartAnalyticsData WareHouseExternal
IBM - IIWIIW is an enterprise-wide data warehousing solution for the insuranceindustryIIW is engineered to consolidate data from disparate systemsSome characteristics- Predefined data models & templates- Based on core concepts- Coverage from requirements to database design- Implements traceability- Supports iterative development approach
Core concepts (packages)The foundation models cover all insurance conceptsafter 6 pmContact preferenceCommunicationContact PointLegal ActionPlaceClaimPartyAgreementActivityMoney ProvisionOOM MMMProductEventPhysical ObjectStandard undAccount
Overview of Industry ModelsTeradata Communications Logical Data ModelTeradata Financial Services Logical Data ModelTeradata Healthcare Logical Data ModelTeradata Insurance Logical Data ModelTeradata Manufacturing Logical Data ModelTeradata Media Logical Data ModelTeradata Retail Logical Data ModelTeradata Transportation and Logistics Logical Data ModelTeradata Travel and Hospitality Industry Logical Data ModelTeradata Utilities Logical Data Model
High level architectureOverview of main functional componentsSource SystemsReportingDatamartAnalyticsData WareHouseExternal
Overview of Industry ModelsAdvanced Analytics (in combination with Detailed Data Store).Risk Management for InsuranceAnti-money Laundering in Financial ServicesExpediting drugs in Life SciencesCross-sell opportunities in retailProducing demand-driven forecasts in manufacturing
High level architectureOverview of main functional componentsSource SystemsReportingDatamartAnalyticsData WareHouseExternal
Insurance Analytics Architecture (IAA) Insurance Detail Data Store (DDS) Subject model High-level, subject area diagram. Logical model Mid-level, business structure of subjects, entities & relationships. Physical model Reflects actual implementation in terms of database and storage system,optimized for performance objectives, etc. Model Metadata Connecting the data model to source data, ETL processes and datamarts
Data Categories in the DDS (part 1)PersonalPropertyShared alPropertyMarketing CampaignPolicyCommercial AutoFinancialAccountPersonalAutoClaims24
Data Categories in the DDS (part 2)CounterpartiesRisk FactorsRisk MitigantsPropertiesFinancial PositionsFinancial FundsMarket DataFinancial AccountsFinancial InstrumentsFinancial InstrumentCalculation Methods
Overview of Industry ModelsOracle AppsProcurement and Spend AnalyticsFinancial AnalyticsSupply Chain and Order Management AnalyticsHuman Resources AnalyticsSales AnalyticsContact Center Telephony Analytics
High level architectureOverview of main functional componentsSource SystemsReportingDatamartAnalyticsData WareHouseExternal
AgendaIntroductionHigh level architectureTechnical AspectsFunctional AspectsOverview of Industry Models (Including Mapping on High level architecture)IBMTeradataSASOracleTechnical aspects of Industry ModelsBest Practices
Technology concepts – Entity informationValidityBusiness dateValid fromValid toBusiness Key(Alternate Key)Source Attributes‘Other Attributes’
Technology concepts – Entity informationGenerated key (Identity)ValidityValid fromValid toBusiness Key(Alternate Key)‘Other Attributes’
Technology concepts – Entity informationGenerated key (Version)Generated key (Identity)ValidityValid fromValid toBusiness Key(Alternate Key)‘Other Attributes’
Technology concepts – Data VaultDV-‘Hub’DV-‘Satellite’Generated key (Version)Generated key (Identity)ValidityValid fromValid toBusiness Key(Alternate Key)‘Other Attributes’
Technology concepts – IIWIIW ‘Anchor’IIW ‘Leaf’Generated key (Version)Generated key (Identity)ValidityValid fromValid toBusiness Key(Alternate Key)‘Other Attributes’
Technology concepts – DDSSAS ‘Entities’Generated key (Version)Generated key (Identity)ValidityValid fromValid toBusiness Key(Alternate Key)‘Other Attributes’
Best Practices1.Successful implementations have both attention for Technical aspects andFunctional aspects of the BI – modelsBy a ‘Technical Approach’ the business alignment is a primary attention point.By a ‘Functional Approach’ the standardization, maintainability and extensibility is a primaryattention point.2. Industry Models are the result of a growth process for years. Be aware of‘legacy’ in these models.There might be legacy constructions that would be implemented in the model in a different wayas they were added during the last years.3. A Choice for an Industry Model, does not mean you have to do an exactimplementation. Using it as a reference model is a valid alternative.4. Modeling Techniques, like Data Vault and Anchor Modeling, will haveadditional value when business content is added.
Technical Aspects Functional Aspects. Overview of Industry Models (Including Mapping on High level architecture) IBM Teradata SAS Oracle. Technical aspects of Industry Models Best Practices. 4. Maarten Ketelaars Managing Consultant 06-18537914 Maarten.Ketelaars@nippur.nl. Professional Background and Skills Employers.
