Data Modelling Techniques for Better Business Intelligence A Focus on the Data Modelling Process Introducing MIKE 2. 0 An Open Source Methodology for Information Development http: //www. openmethodology. org
Better Business Intelligence through an Information Development Approach Agenda n Better Business Intelligence ─ The Keys to Better Business Intelligence ─ Guiding Principles for Better Business Intelligence n MIKE 2. 0 Methodology: A Focus on Data Modelling ─ 5 phased-approach to Better Business Intelligence ─ Example Task Outputs from Strategy Activities ─ Example Task Outputs from Implementation Activities n Lessons Learned © 2008 Bearing. Point, Inc. CROSS 2
Business Intelligence Defining Better Business Intelligence refers to a set of techniques and technologies that are used to gather information from a repository through reports and analytical tools. n Reporting and Analytics can be considered the "front end" of the Business Intelligence environment n Reporting and analytics involve a combination of automated and user-driven steps Business Intelligence typically involves accessing repositories where data is brought together from many different systems across the organisation. n Can be considered the "back-end" of the Business Intelligence environment n The back-end is generally an automated process The delivery approach for Business Intelligence projects is different that functional or infrastructure-related projects. n Seen as more of a "journey" than more functionally-oriented development – The focus is on incremental delivery n Testing can be challenging – It is inherently more difficult to simulate all user cases © 2008 Bearing. Point, Inc. CROSS 3
Business Intelligence Defining Better Business Intelligence In the past many Business Intelligence initiatives have failed: n Most failures were typically due to the "back-end" or the SDLC process n Organisations want a better Business Intelligence environment more than ever – and the capabilities they need today are even more sophisticated Back-end issues have primarily been related to: n Data Integration n Metadata Management n Data Quality Management n Data Modelling Delivery approach issues were primarily related to: n Lack of a strategic vision that allowed for incremental delivery n Poorly defined requirements n Inadequate testing n Architectural flexibility In order to move to a reliable and effective Business Intelligence environment, the focus must be on getting these areas right and taking an Information Development approach. © 2008 Bearing. Point, Inc. CROSS 4
The MIKE 2. 0 Methodology An Open Source Methodology for Information Development Key Constructs within MIKE 2. 0 n SAFE (Strategic Architecture for the Federated Enterprise) is the architecture framework for the MIKE 2. 0 Methodology n Information Development is the key conceptual construct for MIKE 2. 0 – Develop your information like you do with applications n The Overall Implementation Guide provides the overall set of Activities and Tasks that bring everything together; the Usage Model determines what is used depending on the type of project n Supporting Assets are detailed artifacts that link to Activities. Supporting Assets include: ─ ─ Software Assets ─ Deliverable Templates ─ Sales assets ─ n Tools and Technique papers Open Source Examples: http: //mike 2. openmethodology. org/index. php/MIKE 2: Supporting_Assets MIKE 2. 0 Solutions tie Supporting Assets to the Overall Implementation Guide ─ Technology Backplane Solutions are technically-oriented (e. g. MIKE 2. 0 for Business Intelligence) ─ More Solutions are under development, including Business Solutions and Vendor Solutions n MIKE 2. 0 recommends a new organizational model and governance standards to deliver the Information Development Centre of Excellence n Many Private MIKE 2. 0 Assets are stored on internal Bearing. Point content management systems © 2008 Bearing. Point, Inc. CROSS 5
The MIKE 2. 0 Methodology An Open Source Methodology for Information Development MIKE 2. 0 provides a Comprehensive, Modern Approach n Scope covers Enterprise Information Management, but goes into detail in areas to be used for more tactical projects n Architecturally-driven approach that starts at the strategic conceptual level, goes to solution architecture n A comprehensive approach to Data Governance, Architecture and strategic Information Management MIKE 2. 0 provides a Collaborative, Open Source Methodology for Information Development n Balances adding dynamic new content with release stability through a method that is easier to navigate and understand n Allows non-Bearing. Point users to contribute n Links into Bearing. Point's existing project assets on Intraspect n Unique approach, we would like to make this "the standard" in the new area of Information Development © 2008 Bearing. Point, Inc. CROSS 6
The MIKE 2. 0 Methodology Key Activities for Data Modelling Validate Strategic Business Requirements Refine Strategic Business Requirements to Detailed Requirements Categorise Detailed Business Requirements Prioritise Detailed Business Requirements Determine Detailed Analytical Requirements Party Arrangement (Accounts, etc. , ) Campaign Features Organization Product (People/Org. of Interest & their Relationship) Location Event (Content/TXN, etc. , ) Channel (ATM, Kiosk, etc. , ) © 2008 Bearing. Point, Inc. CROSS 7
MIKE 2. 0 Methodology: Phase Overview The 5 Phases of MIKE 2. 0 Information Development through the 5 Phases of MIKE 2. 0 Continuous Implementation Phases Increment 3 Strategic Programme Blueprint is done once Increment 2 Increment 1 Design Phase 1 Business Assessment Phase 2 Technology Assessment Development Roadmap & Foundation Activities Deploy Operate Begin Next Increment Phase 3, 4, 5 Improved Governance and Operating Model © 2008 Bearing. Point, Inc. CROSS 8
MIKE 2. 0 Methodology: Phase Overview Typical Activities Conducted as part of the Strategy Phases Phase 1 – Business Assessment and Strategy Definition Blueprint 1. 1 Strategic Mobilisation 1. 2 Enterprise Information Management Awareness 1. 3 Overall Business Strategy for Information Development 1. 4 Organisational Quick. Scan for Information Development 1. 5 Future State Vision for Information Management 1. 6 Data Governance Sponsorship and Scope 1. 7 Initial Data Governance Organisation 1. 8 Business Blueprint Completion 1. 9 Programme Review Phase 2 – Technology Assessment and Selection Blueprint 2. 1 Strategic Requirements for BI Application Development 2. 2 Strategic Requirements for Technology Backplane Development 2. 3 Strategic Non-Functional Requirements 2. 4 Current-State Logical Architecture 2. 5 Future-State Logical Architecture and Gap Analysis 2. 6 Future-State Physical Architecture and Vendor Selection 2. 7 Data Governance Policies 2. 8 Data Standards 2. 9 Software Development Lifecycle Preparation 2. 10 Metadata Driven Architecture 2. 11 Technology Blueprint Completion © 2008 Bearing. Point, Inc. CROSS 9
MIKE 2. 0 Methodology: Task Overview Task 1. 3. 2 and 1. 3. 3 Define Strategic CSFs and KPIs Information Development through the 5 Phases of MIKE 2. 0 Continuous Implementation Phases Strategic Programme Blueprint is done once Responsible Status Increment 3 Increment 2 Increment 1 Task 1. 3. 1 Define Strategic Business Vision Task 1. 3. 2 Define Strategic Critical Success Factors (CSFs) Design Phase 1 Business Assessment Activity 1. 3 Overall Business Strategy for Information Development Phase 2 Technology Assessment Task 1. 3. 3 Define Strategic Key Performance Indicators (KPIs) Roadmap & Foundation Activities Deploy Task 1. 3. 4 Define Strategic Success Measures Operate Begin Next Increment Phase 3, 4, 5 Improved Governance and Operating Model Task 1. 3. 5 Define Strategic Change Drivers Task 1. 3. 7 Define High-Level Information Requirements Phase 1 – Business Assessment and Strategy Definition Blueprint 1. 1 Strategic Mobilisation 1. 2 Enterprise Information Management Awareness 1. 3 Overall Business Strategy for Information Development 1. 4 Organisational Quick. Scan for Information Development 1. 5 Future State Vision for Information Management 1. 6 Data Governance Sponsorship and Scope 1. 7 Initial Data Governance Organisation 1. 8 Business Blueprint Completion 1. 9 Programme Review © 2008 Bearing. Point, Inc. CROSS 10
MIKE 2. 0 Methodology: Task Overview Task 1. 3. 2 and 1. 3. 3 Define Strategic CSFs and KPIs Analytical Reporting provides focus to address KPI's which drive the business Linear Quantitative Balanced Score Card What if Analysis Key Performance Indicators (KPI's) Critical Success Factors (CSFs) © 2008 Bearing. Point, Inc. CROSS 11
MIKE 2. 0 Methodology: Task Overview Task 1. 5. 7 Define Future-State Conceptual Data Model Information Development through the 5 Phases of MIKE 2. 0 Continuous Implementation Phases Strategic Programme Blueprint is done once Increment 3 Increment 2 Increment 1 Activity 1. 5 Future-State Vision for Information Management Responsible Status 1. 5. 1 Introduce Leading Business Practices for Information Management 1. 5. 2 Define Future-State Business Alternatives Design Phase 1 Business Assessment Development Phase 2 Technology Assessment Roadmap & Foundation Activities 1. 5. 3 Define Information Management Guiding Principles 1. 5. 4 Define Technology Architecture Guiding Principles Deploy Operate Begin Next Increment Phase 3, 4, 5 1. 5. 5 Define IT Guiding Principles (Technology Backplane Delivery Principles) 1. 5. 6 Define Future-State Information Process Model Improved Governance and Operating Model 1. 5. 7 Define Future-State Conceptual Data Model Phase 1 – Business Assessment and Strategy Definition Blueprint 1. 1 Strategic Mobilisation 1. 2 Enterprise Information Management Awareness 1. 3 Overall Business Strategy for Information Development 1. 4 Organisational Quick. Scan for Information Development 1. 5 Future State Vision for Information Management 1. 6 Data Governance Sponsorship and Scope 1. 7 Initial Data Governance Organisation 1. 8 Business Blueprint Completion 1. 5. 8 Define Future-State Conceptual Architecture 1. 5. 9 Define Source-to-Target Matrix 1. 5. 10 Define High-Level Recommendations for Solution Architecture 1. 9 Programme Review © 2008 Bearing. Point, Inc. CROSS 12
MIKE 2. 0 Methodology: Task Overview Task 1. 5. 7 Define Future-State Conceptual Data Model The Conceptual Model records the broad objects/things (sometimes called 'subject areas') that the business interacts with and names the relationships between these. The purpose of the Conceptual model is to discover the big ticket items and to name them in an agreed way. (People/Org. of Interest & their Relationship) Party Arrangement (Accounts, etc. , ) Campaign Features Organization Product © 2008 Bearing. Point, Inc. Location Event (Content/TXN, etc. , ) Channel (ATM, Kiosk, etc. , ) CROSS 13
MIKE 2. 0 Methodology: Task Overview Task 1. 5. 10 High Level Solution Architecture Options Information Development through the 5 Phases of MIKE 2. 0 Continuous Implementation Phases Strategic Programme Blueprint is done once Increment 3 Increment 2 Increment 1 Activity 1. 5 Future-State Vision for Information Management Responsible Status 1. 5. 1 Introduce Leading Business Practices for Information Management 1. 5. 2 Define Future-State Business Alternatives Design Phase 1 Business Assessment Development Phase 2 Technology Assessment Roadmap & Foundation Activities 1. 5. 3 Define Information Management Guiding Principles 1. 5. 4 Define Technology Architecture Guiding Principles Deploy Operate Begin Next Increment Phase 3, 4, 5 1. 5. 5 Define IT Guiding Principles (Technology Backplane Delivery Principles) 1. 5. 6 Define Future-State Information Process Model Improved Governance and Operating Model 1. 5. 7 Define Future-State Conceptual Data Model Phase 1 – Business Assessment and Strategy Definition Blueprint 1. 1 Strategic Mobilisation 1. 2 Enterprise Information Management Awareness 1. 3 Overall Business Strategy for Information Development 1. 4 Organisational Quick. Scan for Information Development 1. 5 Future State Vision for Information Management 1. 6 Data Governance Sponsorship and Scope 1. 7 Initial Data Governance Organisation 1. 8 Business Blueprint Completion 1. 5. 8 Define Future-State Conceptual Architecture 1. 5. 9 Define Source-to-Target Matrix 1. 5. 10 Define High-Level Recommendations for Solution Architecture 1. 9 Programme Review © 2008 Bearing. Point, Inc. CROSS 14
MIKE 2. 0 Methodology: Task Overview Task 1. 5. 10 High Level Solution Architecture Options Show below are sample outputs of high-level Solution Architecture options at the level they would be produced for this task. Typically, there will be a few architecture models with supporting text. This proposed solution includes 3 viable options: n n n Use a Vendor model as the base logical data model for integrated Operational Data Store, going through a map-and-gap exercise to complete the model. This model is closely aligned to the existing data classification/taxonomy model that has been adopted organisation-wide Develop & build a hybrid data model consisting of existing data models used across the organisation from existing systems. These base models will need to be supplemented and integrated with other models currently used in enterprise applications Develop and build a logical, normalised data model in-house for the, based on the existing data classification/taxonomy model that has been adopted organisation-wide and a well-defined set of user requirements © 2008 Bearing. Point, Inc. Reference Model Option 1 Vendor Model * Contract admin * CRM * Pricing Systems Option 2 System XXX System YYY *Product systems Reference Model Option 3 In-house CROSS 15
MIKE 2. 0 Task Overview: Task Overview Task 2. 11. 3 Define Capability Deployment Timeline Information Development through the 5 Phases of MIKE 2. 0 Continuous Implementation Phases Strategic Programme Blueprint is done once Increment 3 Increment 2 Activity 2. 11 Technology Blueprint Completion Responsible Status Task 2. 11. 1 Revise Blueprint Architecture Models Increment 1 Task 2. 11. 2 Define Major Technology Risks and Constraints Design Phase 1 Business Assessment Phase 2 Technology Assessment Task 2. 11. 3 Define Business and Technology Capability Deployment Timeline Development Roadmap & Foundation Activities Deploy Operate Begin Next Increment Task 2. 11. 4 Revise Business Case Phase 3, 4, 5 Task 2. 11. 5 Define Roadmap Mission Statements Task 2. 11. 6 Assemble Key Messages to Complete Technology Blueprint Improved Governance and Operating Model Phase 2 – Technology Assessment and Selection Blueprint 2. 1 Strategic Requirements for BI Application Development 2. 2 Strategic Requirements for Technology Backplane Development 2. 3 Strategic Non-Functional Requirements 2. 4 Current-State Logical Architecture 2. 5 Future-State Logical Architecture and Gap Analysis 2. 6 Future-State Physical Architecture and Vendor Selection 2. 7 Data Governance Policies 2. 8 Data Standards 2. 9 Software Development Lifecycle Preparation 2. 10 Metadata Driven Architecture 2. 11 Technology Blueprint Completion © 2008 Bearing. Point, Inc. CROSS 16
MIKE 2. 0 Task Overview: Task Overview Task 2. 11. 3 Define Capability Deployment Timeline Prod 1 Data Model Prod 2 Data Model Integrated Data Model First Increment Cust Analysis n Enterprise Wide Stakeholders Community definition with roles and responsibilities n First Enterprise Wide Enterprise Warehousing Workshop Integrate n Prod 1 Source System Attribute Selection Prod 2 Source System Attribute Selection Functional Capabilities of a comprehensive ODS, Warehouse and Data Mart environment Second Increment n Completed Analysis on the availability of sources for cost information (e. g. , atomic data and Cross-Over Tables) n Source Implementations of (e. g. , atomic data and Cross-Over Tables) for cost information n Completed Analysis for Customer and Product Profitability Analysis n n Completed Analysis on all Cross Sectional generating events. Initial implementations for Customer and Product Profitability Analysis n n Completed 'Whole of Customer' matching strategy across Households and Products Metadata management applications extended to a limited user 'self service' environment n Messaging and Real-Time Info-Structure implemented for initial round of ODS, Warehouse and Mart access Full Scale Sourcing Prod 1 Full Scale Sourcing Prod 2 Initial Sourcing Info-Structure/ODBC Integration with ODS/Warehouse Integrated ODS Warehouse Production Implementation Iterative Application Integration Full Sourcing Integrated Metadata Management Course Correction from Partial ODS/Warehouse n Customer and Product ODS implementation n AR closed loop to the warehouse designed n Finance and Service information designed for incorporation in the EDW n Proprietary environment used as a Data Mart n Ongoing Data Quality Monitoring in place n EDW development and management organization established n EDW contains base information for accounts, customers and products n Detail Integrated Program of Works n n Detail Integration Methodology and implementation process Production use of the initial data warehouse implementation n n Initial Integrated Data Model Full Scale Sourcing for multiple retail products Initial Integrated Metadata Model n Initial Sourcing for customers and products Enterprise Wide Standards for attribute models, message models n Second phase of Household matching and first phase of product matching n Functional requirements for the warehousing Info-Structure n Meta. Data repository available in production environment Initial Data Schemas allocated in a physical environment n An ongoing leader of enterprise information established Initial Source systems identified for initial attributes n n Business Rules for all data cleansing identified Second enterprise wide workshop on data warehousing is held n Integrate Enterprise Priorities mapped to the Functional Capabilities n Cust Design n n Initial Warehouse Implementation Revenue/Whole of Customer Continuing Analysis Tasks identified n Initial Warehouse operational for testing and validation n n Third Increment n n First EIS dashboard based upon the Enterprise Data Warehouse deployed The second release of the decision support models for DSS © 2008 Bearing. Point, Inc. CROSS 17
MIKE 2. 0 Task Overview: Task Overview Task 2. 11. 3 Define Capability Deployment Timeline Whole of Customer Revenue View – The focus of this component is on bringing together the 'Whole of Customer' for Product 1 and Product 2 from the perspective of Revenue. Initial matching of customers will begin; however, this will not limit product operational systems from using the information from their own perspectives. Whole of Product Revenue View – The focus of this component is to begin the 'Whole of Product" view. The revenue information comes from XXXXX (source: XXXX) and XXXX. Product revenue will be tracked by the current segmentation in these systems as well as the product structures in these systems. Complex Customer/Product Formulation – The focus of this effort will be to formulate some of the more complex definitions of customer and product. These activities, initially, will perform the required customer and product business analysis to enhance the warehouse data models. Cross-Sectional Formulations – The focus of these efforts will be to establish the initial understandings of how the warehouse information must be summarized. Examples are: week, month, quarter, year, identified customer or product event. Dependent Data Mart Formulation – The Dependent Data Marts addressed the specific business support needs of particular Enterprise business constituencies. The Marts can contain historical as well as ODS information. They will be used for a number of activities such as reporting or query as well as analytical activities. Decommissioning – This thread of activities will focus on the decommissioning of the current high maintenance ODS/MIS implementations. The XXXXXXX, XXXXX and XXXXXX Databases are key in the decommissioning process. Unneeded capabilities can be terminated while others are target for the new environment. Common Info-Structure – This effort focuses on the hardware and network environment for the implementation and use of the Enterprise Data Warehouse Environment. ETL and EAI implementations will be key. The hardware options will address ODS, Warehouse and Mart Environments. First Increment Common Data Model Second Increment Prod 1 Customer Revenue Load Local Customer Revenue Load Common Data Model Initial Use of Prod 1 Info Initial Use of Local Info Product Taxonomy Product X Taxonomy Extended Customer Definitions Customer Matching across X and Y Products Product Aggregates Product 3 Product Revenue ODS and Mart Implementations Product Revenue to Projects Analysis Taxonomy of Customer Profiles New Product Models Mapping of Customer and Product Profiles Taxonomy of Product Profiles EIS Decision Models Monthly Weekly Customer Revenue ODS and Mart Implementations Product Y Summary Extended Product Definitions Daily Third Increment Yearly EIS Dashboards DSS Information Support DSS Decision Models Event Driven Mart Constituency Inventory Mart Constituency Requirements ODS Support Data Mart Models and Tools Initial Data Mart Implementation Current ODS/MIS Users Inventory Functions to Migrate Inventory Decommissioning and Discontinuing Current ODS/MIS Function Inventory Functions to Discontinue Inventory DB Hardware Implementation ETL and Warehouse Tools Implemented © 2008 Bearing. Point, Inc. Ongoing Data Quality Improvement SOA/Info-Structure and Security Implementation CROSS 18
MIKE 2. 0 Methodology: Phase Overview Roadmap and Foundation Activities The MIKE 2. 0 Roadmap covers the planning, requirements and conceptual design for each increment. Foundation Activities are what we want to get out "in front" in the information management imitative. Phase 3 – Information Management Roadmap and Foundation Activities 3. 1 Information Management Roadmap Overview 3. 2 Testing and Deployment Plan 3. 3 Software Development Readiness 3. 4 Detailed Release Requirements 3. 5 Business Scope for Improved Data Governance 3. 6 Enterprise Information Architecture 3. 7 Root Cause Analysis on Data Governance Issues 3. 8 Data Governance Metrics 3. 9 Database Design 3. 10 Message Modelling 3. 11 Data Profiling 3. 12 Data Re-Engineering 3. 13 Business Intelligence Initial Design and Prototype 3. 14 Solution Architecture Definition/Revision © 2008 Bearing. Point, Inc. CROSS 19
The MIKE 2. 0 Methodology Activity 3. 4 Detailed Release Requirements Information Development through the 5 Phases of MIKE 2. 0 Continuous Implementation Phases Strategic Programme Blueprint is done once Increment 3 Increment 2 Increment 1 Phase 2 Technology Assessment Task 3. 4. 1 Validate Strategic Business Requirements Task 3. 4. 2 Refine Strategic Business Requirements to Detailed Requirements Design Phase 1 Business Assessment Activity 3. 4 Detailed Business Requirements Development Roadmap & Foundation Activities Task 3. 4. 3 Categorise Detailed Business Requirements Deploy Operate Begin Next Increment Phase 3, 4, 5 Task 3. 4. 4 Prioritise Detailed Business Requirements Task 3. 4. 5 Determine Detailed Analytical Requirements Improved Governance and Operating Model Phase 3 – Information Management Roadmap and Foundation Activities 3. 1 Information Management Roadmap Overview 3. 2 Testing and Deployment Plan 3. 3 Software Development Readiness 3. 4 Detailed Release Requirements 3. 5 Business Scope for Improved Data Governance 3. 6 Enterprise Information Architecture 3. 7 Root Cause Analysis on Data Governance Issues 3. 8 Data Governance Metrics 3. 9 Database Design 3. 10 Message Modelling 3. 11 Data Profiling 3. 12 Data Re-Engineering 3. 13 Business Intelligence Initial Design and Prototype 3. 14 Solution Architecture Definition/Revision © 2008 Bearing. Point, Inc. CROSS 20
The MIKE 2. 0 Methodology Activity 3. 4 Detailed Release Requirements Validate Strategic Business Requirements Refine Strategic Business Requirements to Detailed Requirements Categorise Detailed Business Requirements Prioritise Detailed Business Requirements Determine Detailed Analytical Requirements Party Arrangement (Accounts, etc. , ) Campaign Features Organization Product (People/Org. of Interest & their Relationship) Location Event (Content/TXN, etc. , ) Channel (ATM, Kiosk, etc. , ) © 2008 Bearing. Point, Inc. CROSS 21
The MIKE 2. 0 Methodology Task 3. 9. 1 Develop Logical Data Model Information Development through the 5 Phases of MIKE 2. 0 Continuous Implementation Phases Strategic Programme Blueprint is done once Increment 3 Activity 3. 9 Database Design Task 3. 9. 1 Design Logical Data Model Increment 2 Task 3. 9. 2 Develop Physical Data Model Increment 1 Design Phase 1 Business Assessment Phase 2 Technology Assessment Development Roadmap & Foundation Activities Deploy Operate Begin Next Increment Phase 3, 4, 5 Improved Governance and Operating Model Phase 3 – Information Management Roadmap and Foundation Activities 3. 1 Information Management Roadmap Overview 3. 2 Testing and Deployment Plan 3. 3 Software Development Readiness 3. 4 Detailed Release Requirements 3. 5 Business Scope for Improved Data Governance 3. 6 Enterprise Information Architecture 3. 7 Root Cause Analysis on Data Governance Issues 3. 8 Data Governance Metrics 3. 9 Database Design 3. 10 Message Modelling 3. 11 Data Profiling 3. 12 Data Re-Engineering 3. 13 Business Intelligence Initial Design and Prototype 3. 14 Solution Architecture Definition/Revision © 2008 Bearing. Point, Inc. CROSS 22
The MIKE 2. 0 Methodology Task 3. 9. 1 Develop Logical Data Model The Logical Data Model (LDM) is a more formal representation of the conceptual and contains far greater supporting detail. Relational theory is used to normalise the data, like objects may be grouped into super and sub types, many-to-many relationships are resolve. © 2008 Bearing. Point, Inc. CROSS 23
The MIKE 2. 0 Methodology Supporting Assets – Mapping to an Off the Shelf Data Model Subset of a Supporting Asset: Application CIF Extension In general the design of a Data Mart M (if Dimensional modeling – Star Schema involved), would involve the following methodology approach. n n Identify a business process/requirement (i. e. ALM requirements, MIS reports etc). A DM is designed around "known" requirements Identification of the lowest level of the process (i. e. Individual txn, individual daily/mthly snapshot), which will be represented in the fact table for this process Analyze the elements of the business process or requirements and identify the Dimension, Measure, etc of the process with related characteristics (i. e. hierarchies, aggregates, history, etc) noted For each business process the identification of the Fact, related Dimension tables, their contents and relationships between the tables are pursued. Where there are multiple business process or requirements within a subject area (i. e. ALM, Profitability, etc) this approach will continue Note: If the DM logical design were based on a subset of the FSLDM, a similar process as discussed for how to design a DW would be pursued. © 2008 Bearing. Point, Inc. CIF Facility Account Customer Account Relationship Application Monthly Summary Application Facility Account Application Daily Summary Application Transaction 3 NF General (Common) Dimensions Time Product Sales $ Revenue Volume Geography Facts Color Star Schema Customized (Complex) CROSS 24
Better Business Intelligence Lessons Learned Define a Strategy that can be Executed n Launch a large-scale top-down strategy with a bottom-up (narrow and detailed) engagement if necessary n Always define the tactical within the strategic and plan for re-factoring and continuous improvement in the overall programme plan n Focus on improving key data elements – Don't do everything at once Design a Strategy that is Flexible and Meaningful to the Business n Expect business requirements to change – Provide an infrastructure to handle a dynamic business n Know your risk areas in each implementation increment – Focus on foundation activities first n Be aware of technology lock-in and know the cost of "getting out" – Use an open approach n Break through limiting factors in legacy technology – This is the opportunity to kill the sacred cows Keep the Business Engaged n Communicate continuously on the planned approach defined in the strategy – The overall Blueprint is the communications document for the life of the programme n Always focus on the business case – Even for initial infrastructure initiatives or replacement activities © 2008 Bearing. Point, Inc. CROSS 25
Скачать презентацию Data Modelling Techniques for Better Business Intelligence A
2665c7d66814a7d341d747040a6d3daa.ppt