Transcription

Context is EverythingUsing policies, data quality and stewardship to govern informationAshwin SinhaPractice Manager – Information [email protected]: August 2013www.certussolutions.com

Agenda Information Governance Business Glossary University example Context of Data Quality Data Quality Framework Certus in University Questions

Six Steps to Governance1.Set your Goals - the core statements that guide the operation and development of theinformation supply chain.2.Define Your Metrics - the set of measurements used to assess the ongoing effectiveness ofthe program and associated governance processes.3.Make Decisions - the organisational structure and changing ideological model to analyse andmake policy decisions.4.Communicate Policy - the tools, skills and techniques used to communicate policy decisions tothe organisation.5.Measure Outcomes – Compare policy results with goals, inputs, decision models, andcommunication to provide constant feedback on policy effectiveness.6.Audit results – the tool you use to benchmark everything.

Information Governance – Key areas

Information Governance – sTermsPlanningData QualityScopingSecurity tureGlossaryTrainingDeploy KPIMonitoringMonitor Qualityand cardConnectRules liciesCustomiseIndustryModelPilot MDMProfilingDeployModel toGlossaryReference DataManagementStewardshipand Review

Definition of a Glossary What is a Glossary? Most Glossaries are labels and descriptions stored in PDF, Word or HTMLformat. There is a lack of consistency, ownership and re-use across theseGlossaries. They are expensive to build and maintain.

Glossary for Universities Student ? EFTSL – Equivalent Full Time Student Load. That is, a full timestudent with 24 subject units will have an EFTSL of 1, and a halftime student will have an EFTSL of 0.5 An EFTSL - or Equivalent Full Time Student Load - is arepresentation of the amount of load a student would havewhen studying full time for one year. At Curtin 200 credit pointsequates to 1.0 EFTSL. The measure used to determine a student's enrolled load. One'EFTSL' is the amount of student load determined by theUniversity to be equal to a full-time load for one student forone year, and is expressed at the University as 36 units.

More on same termAt the University of Sydney: 1 EFTSL is equivalent to 48 credit points (1 year of fulltime study)How do I find out the EFTSL of a unit of study? The EFSTL a unit of study can be determined by dividingthe unit credit point value by 48Can we have a consistent meaning of above terms always,in absence of context?

Why Use a Glossary?

Central Nervous System The Central Nervous System receives information from all parts of the bodyand coordinates activity.Data IntegrationPrivacy ta QualityPoliciesGlossaryMaster DataManagementBig Data

Giving Context to Information Under a traditional documentation and development approach artefactsdecline in valueContainGet LostRepresent oldinaccuraciesBecome StaleSomewherestandards In a Metadata Driven architecture Business Glossary and data lineagebecomes more valuable over time.Retains SMEKnowledgeShared acrossprojectsReflects the realworldDrives ChangeManagement

DQM – Why?Benefits of undertaking a dedicated data quality management (DQM)initiative as part of the Information Management programme include:– Engagement on DQM as a strategic issue throughout the enterprise– Cost-effective actions, targeted at the most critical issue– Supports treating information as a strategic asset that supports business processexecution– Measurable outcomes. Providing context to Data Quality issues.– Sustainability for the long term– Assurance that business value & benefit meets the initial business case– Clear roadmap for future progress

Data Quality Framework Mobilising People, Processes and Tools to improve the quality ofinformationInformationDataProfilingAnalyzerAnd AnalysisBusiness GlossaryPoliciesdefineobjectivesreportassess &discoverData QualityScorecardvalidatecleanse &enrichmonitor /track“master”Data TransformationException ReportingData Rule andExceptionManagementData CleansingAnd Mastering

Improving Data Quality SafeguardsData Quality Scorecards are part of a framework to reduce risk: The Data Quality Scorecard is a measure of problems, the overall Information Management framework of Business Glossary,Metadata reporting and exception handling provides the support for handling these problems. A Scorecard is the ongoing audit that data is safe to use and the safeguard against complacency.Business Glossary:what do we knowabout this data?Stewardship: who isresponsible for thisdata?Data Lineage: wheredid the problemoccur?Exception Management:what do we do aboutthis problem?1

DQM – How (Scalable Approach)ScopeMeasureAction Define DQM andboundaries Establish DQ analysisenvironment Establish DQ resolutionenvironment Ensure stakeholderbuy-in Profile data Address DQ issues People Process Technology Capture business rules Define data domains Root cause analysis Identify DQ Issues Quantify businessimpacts Qualitative Quantitative Agree action plan Establish DQprevention activity People Process TechnologyOperate Embed DQM as BAU Extend and evolveDQM acrossorganisation Extend and evolve aspart of approach toInformationManagement

DQM – How (Scalable Approach)

DQM – Keys to SuccessThere are many factors to ensure the success of establishing andoperating a DQM capability:– DQM needs to be clearly linked to business objectives with business sponsorship– Clearly articulated business impacts both quantitative as well as qualitative– Clearly understand root causes and costs to remediate / prevent People, Process and Technology– Start with high value and low cost to demonstrate value and ensure ‘right-sized’approach– Define KPI’s that leverage D.A.T.A. Digestible,ActionableTimelyAuditable

Certus provides the frameworkWhat rule sets are available?A combined set of over 250 data quality rules and over 200 Business Glossary termsfor Education sector developed and provided by Certus and IBM is available to kickstart a project.Examples of the rules available include: Party/person rules - validation of an actual person's name, gender and age.Some examples of this are reasonableness checks on whether name is in avalid format, age range is valid, etc. Address rules - validation of the various parts that make up the address, andreference checks of cities and post codes. Email and web rules - validate the format of email addresses, email domainnames, host names, URLs and IP addresses. Field format rules- A large number of generic field format validation rulesthat cover a range of different field types such as indicators, codes, dates,numeric formats, etc.

Certus DQF Overview

Data Quality FrameworkThe artefacts within the Data Quality Framework provide the client with a top downunderstanding of the implementation of the Data Quality Process. The Frameworkgives you an understanding of the roles, responsibilities, tasks and processes involvedin the implementation.The DQF Hierarchy Tree defines the high level steps required for the deployment ofthe framework. These steps then break down into supporting artefacts at the medianlevel, providing the client with:– Business Process Diagram for each main level in the Hierarchy tree.– RACI document which details the tasks and technologies involved in eachlevel of the hierarchy tree and roles and responsibilities across each task.– Project Plan templateTechnical artefacts are also included that detail each step involved in the process.

Certus DQF Framework

DQF Framework Components Level 1CDQF Hierarchy Tree - Top level processes involved in the framework. Level 2Detailed artefacts supporting the top level processes RACI DocumentRoles – An outline of the roles required within your organisation for this initiative.Responsibilities – responsibilities outlined and assigned by role.Technologies – technology components required in each step. Project PlanThe project plan provides a template for deployment covering each task withinthe Framework broken down by duration and role involved. Business Process Method MapProvides a visual representation of the flow of tasks throughout each businessprocess.

DQF Technical ComponentsDQF Data Model The DQF data model provides the underlying store for the metadata that istransferred throughout the framework.Framework Integration ComponentsThe components below make up the technical plumbing of the Framework. ETL (IBM DataStage) Database Scripts (DB2, Oracle, SQL Server) Operating System Script (Linux, Windows)Certus DQ Dashboard and Active Reporting Solution High Level Dashboard Reporting that gives you the ability to drill down into the high level metrics. Available for iPad for mobilityData Rule Sets Data rules sets are available to assist in the development of your own internal datarules. They give you a head start with common rules found in separate data setsi.e. Name and Address.

Educational ExperienceWho we’ve been working with.Two examplesJames Cook UniversityNSW Board of StudiesChristchurch PolytechnicInstitute of TechnologyEdith Cowan UniversityDeakin UniversityGriffith UniversityMassey UniversityMinistry of EducationNSW Department of Education & CommunitiesNSW TAFEOpen UniversityStudy Group InternationalSouthern Cross UniversityUniversity of AucklandUniversity of South AustraliaUniversity of Technology SydneyUniversity of WollongongJames Cook UniversityAs part of Information to Analytics projectCertus team has worked with JCU to establishan information management framework, BIreporting and Budgeting & Planning. Solution. Italso addresses some of the early InformationGovernance requirements.Open UniversityThe key development work was on studentenrolments and the cost of running particularcourses. It was enrolment-driven, looking atthe individual subject units, and the studentprofiles which may contribute to additionalcourse requirements.

What we doEnhance operational efficiency andoptimise management processes in yourasset intensive t people, share information andempower your organisation to worksmarterCreate, utilise and maintain trustedinformation assets with IM strategiesand solutionsINFORMATIONMANAGEMENTWEBSOLUTIONSDrive improved business performancewith unique user-centred designmethodologiesDeliver complete, consistent andaccurate information to decision-makersfor improved business be to flexible IBM software supportand access expert technical resources tomaintain operational efficiencyMaintain system performance andunleash productivity with reliable,on-demand business ITECTURE& SDLCAlign corporate strategy with businessprocess, information, applications andtechnologyBuild enterprise applications andremove information silos for improvedbusiness agilityMIDDLEWARESOLUTIONSLICENSINGExtract maximum value from your IBMsoftware with effective, compliantlicence management

Thanks & Questions

FAQ How dependent is the solution on customer having key components of IBM InfoSphere andIBM Cognos? Can the solution be ported to other technologies with a limited effort?To realise the full benefits of the solution the customer would need to be an IBM InformationServer and Cognos user.The framework is also useful to clients who wish to implement a DQ project on separatetechnology. The architectural and project management components give the customer ageneric framework to implement there project.

FAQ How flexible is the DQ rule creation and editing for future maintenance?The data rules that are created for the project are easily edited from the IA interface. What sort of support would be needed to keep the solution running?Once the technical components of the project are productionised a number of roles will haveto be filled to ensure an ongoing data quality improvement. A technical resource to monitorthe execution and deal with issues that arise. A business resource from each area to ensurethe data quality results are monitored and initiatives are taken to ensure the improvement ofthat data over time. What will justify the ROI business case for a customer buying this?In our experience full implementation of the Data Quality Framework would save thecustomer up to 12 months of development effort getting it off the ground from scratch.

FAQ What the pre-requisites on software, hardware and discovery to finish this activity?At a generic level the framework is a tool to manage and plan a data quality project.At the full technical level the Framework requires IBM Information Server Components;IBM Information Analyzer V8.7 minIBM Datastage V8.7 minIBM Business Glossary V8.7 minIBM Cognos User licences V10.1 min How do we ensure that the solution is flexible for customers with different data qualityneeds and different types of data?The framework has been generically designed it provides a structure to implement over anindustry. It is not s rigid structure or formulated from an industry specific point of view.

Six Steps to Governance 1. Set your Goals - the core statements that guide the operation and development of the information supply chain. 2. Define Your Metrics - the set of measurements used to assess the ongoing effectiveness of the program and associated governance processes. 3. Make Decisions - the organisational structure and changing ideological model to analyse and