‏‏‏‏‏‏‏‏Chapter 11: Meta-data Management

Meta-data Management
Ahmed Alorage

Objectives:
• 11.1 Introduction
• 11.2 Concepts and Activities
• 11.2.1 Meta-data Definition
• 11.2.1.1 Types of Meta-data
• 11.2.1.2 Meta-data for Unstructured Data
• 11.2.1.3 Sources of Meta-data
• 11.2.2 Meta-data History 1990 - 2008
• 11.2.3 Meta-data Strategy
• 11.2.4 Meta-data Management Activities
• 11.2.4.1 Understand Meta-data Requirements
• 11.2.4.1.1 Business User Requirements
• To be continue…

Objectives:
• 11.2.4.2 Define the Meta-data Architecture
• 11.2.4.2.1 Centralized Meta-data Architecture
• 11.2.4.2.2 Distributed Meta-data Architecture
• 11.2.4.2.3 Hybrid Meta-data Architecture
• 11.2.4.3 Meta-data Standards Types
• 11.2.4.3.1 Industry / Consensus Meta-data standards
• 11.2.4.3.2 International Meta-data standards
• 11.2.4.4 Standard Meta-data Metrics
• 11.2.4.5 Implement a Managed Meta-data Environment
• 11.2.4.6 Create and Maintain Meta-data
• 11.2.4.7 Integrate Meta-data
• 11.2.4.8 Manage Meta-data Repositories
• 11.2.4.8.1 Meta-data Repositories
• 11.2.4.8.2 Directories, Glossaries and Other Meta-data Stores
• 11.2.4.9 Distribute and Deliver Meta-data
• 11.2.4.10 Query, Report and Analyze Meta-data

11 Meta-data Management
• Meta-data Management is the ninth Data Management
Function in the Data Management framework in Chapter
1.
• eighth data management function that interacts with
and influenced by Data Governance function.
• In this Chapter, we will define the meta-data
Management Function and Explains the Concepts and
Activities involved.

11.1 Introduction:
• Meta-data “data about data”:
• Data reflects real life transactions, events, objects, relationships, etc.
• Meta-data reflects data transactions, events, objects, relationships, etc.
• Meta-data Management is:
• the set of processes that ensure proper creation, storage, integration, and
control to support associated of meta-data.
• To understand Meta-data role in Data Management, Example:
• Card catalog in a library who identified the book information in library.
Without this card. It difficult of reader to find the right book, and incorrect
search.
• Figure 11.1, is a short-hand representation of the meta-data
management functions. “the Context diagram”
• Meta-data management activities are in the center, surrounded by
the relevant environmental aspects.

11.1 Introduction:
• Benefits of leveraging meta-data in organization:
• Increase the value of strategic information (e.g. DW, CRM, SCM, etc.) to
providing a context for the data and aiding analysts in making effective
decisions.
• Reduce training costs and lower impact of staff through documentation of
data context, history, and origin.
• Reduce data-oriented research time by assisting business analysts in finding
the information they need, in a timely manner.
• Improve the bridge between business users and IT professional, leverage
work done by other teams, and increase confidence in IT system data.
• Increase speed of system development’s time-to-market by reducing system
development life-cycle time.
• Reduce risk of project failure through better impact analysis at various levels
during change management.
• Identify and reduce redundant data and processes, thereby reducing rework
and use of redundant, out-of-date, or incorrect data.

‏‏‏‏‏‏‏‏Chapter 11: Meta-data Management

11.2 Concepts and Activities
• Meta-data is the card catalog in managed data environment.
• The descriptive tags or context on the data (the content) in a
managed data environment.
• Show business and technical users where to find information in
data repositories.
• Provide details on:
• Where the data came from?
• How has it got there?
• Any transformations
• Its level of quality
• Provide assistance with what the data really means and how to
interpret it.

11.2.1 Meta-data Definition
• It is information about the physical data, technical and business
processes, data rules and constraints, and logical and physical
structures of the data, used by organization.
• These descriptive tags describe data:
• Databases
• Data elements
• Data models
• Concepts:
• Business Processes
• Application systems
• Software code
• Technology infrastructure
• Connections (relationships): between the data and concepts.
• It is broad concept includes many potential subject areas.

11.2.1 Meta-data Definition
• 18 potential subject areas under meta-data management:
1. Business Analytics: Data definitions, reports, users, usage, performance
2. Business Architecture: Roles & organizations, goals and objectives
3. Business definitions: terms for specific concepts, facts in organization
4. Business rules: standard calculations and derivation methods.
5. Data Governance: policies, standards, procedures..etc..
6. Data integration: Sources, targets, transformations, lineage, ETL workflows, EAI,
EII, migration/ conversion.
7. Data quality: Defects, metrics, ratings
8. Document and Content management: Un-data, documents, taxonomies..etc..
9. IT infrastructure: platforms, networks, configurations, licenses.
10. Logical data models: Entities, attributes, relationships and rules..etc..
11. Physical data models: files, tables, columns, views…etc.
12. Process models: functions, activities, roles, inputs/outputs..etc..
13. Systems portfolio and IT governance: DBs, Apps, projects..etc..
14. SOA “Service Oriented Architecture” information: components, services..etc.
15. System design and development: Requirements, designs..etc.
16. Systems Management: Data security, licenses, configuration, service levels.Ect.

Meta-data Definition
11.2.1.1 Types of Meta-data
• Classified into four major types:
• Business, Technical & operational, Process, and data Stewardship
• Business meta-data examples:
• Business data definitions, including calculations.
• Business rules and algorithms, including hierarchies.
• Data lineage and impact analysis.
• Data model: enterprise level conceptual and logical
• Data Quality statements, such confidence and completeness indicators.
• Data stewardship information and owning organization(s)
• Data update cycle
• Historical data availability
• Historical or alternate business definitions.
• Regulatory or contractual constraints.
• Reports lists and data contents.
• System of record for data elements
• Valid value constraints ( sample or list)

• Technical meta-data include DB table, column name, properties..etc.
• Operational meta-data include information about data movement,
source and target systems, batch programs..etc.
• Technical & Operational meta-data examples:
• Audit controls and balancing information
• Data archiving and retention rules
• Encoding/ reference table conversions
• History of extracts and results
• Identification of source system fields
• Mapping, transformation, statistics from (OLTP, OLAP)
• Physical data model, including table names, keys, and indexes
• Program job dependencies and schedule.
• Program names and descriptions
• Purge criteria
• Recovery and backup rules
• Relationships between the data models and DW/ marts
• Systems of record feeding target data stores(OLTP, OLAP, SOA)
• User report and query access patterns, frequency, and execution time.
• Version maintenance

• Process meta-data is data that defines and describes the
characteristics of other system elements (processes, business rules,
jobs, tools, etc.)
• Process meta-data examples:
• Data Stores and data involved
• Government / regulatory bodies
• Organization owners and stakeholders
• Process dependencies and decomposition
• Process feedback loop documentation
• Process name
• Process order and timing
• Process variations due to input or timing
• Roles and responsibilities
• Value chain activities

• Data stewardship meta-data is data about data stewards,
stewardships processes, and responsibility assignments. “assure
data and meta-data accurate, high quality across enterprise”
• Data stewardship meta-data examples:
• Business drivers / goals
• Data CRUD rules
• Data definitions – business and technical
• Data owners
• Data sharing rules and agreements / contracts
• Data stewards, roles and responsibilities.
• Data stores and systems involved
• Data Subject areas.
• Data users
• Government / regulatory bodies
• Governance organization structure and responsibilities

11.2.1.2 Meta-data for Unstructured Data
• There is nothing his name is “Unstructured meta-data”, the better
term “meta-data for unstructured data”.
• Example of meta-data repositories describing unstructured data:
• Content management applications, university websites..etc.
• Meta-data for unstructured meta-data are:
• Descriptive, Structural, or administrative meta-data.
• Descriptive meta-data: (Catalog information, Thesauri keyword terms)
• Structural meta-data: (Dublin Core, Field structures, Format ”
Audio/ Visual, booklet”, Thesauri keyword labels, XML schemas).
• Administrative meta-data (Sources, Integration/update schedule,
Access rights, Page relationships “e.g., site navigational design”).
• Meta-data schemes:
• Bibliographic meta-data: library card of the document.
• Record-keeping meta-data: concerned with validity and retention.
• Preservation meta-data: concerned with storage, archival condition, and
conservation of material.

11.2.1.3 Sources of Meta-data
• Meta-data is everywhere in every data management activity.
• The identification of any data is meta-data of potential interest to
some user group.
• Create business meta-data through user interaction, definition, and
analysis of data.
• Identify meta-data at:
• Aggregate ( such as subject are, system characteristic).
• Detailed ( such as DB column characteristic, code value) level.
• Proper management and navigation between related meta-data is
an important usage requirement.
• Sources of meta-data:
• Primary sources of meta-data are numerous – virtually anything named in an
organization.
• Secondary sources are other meta-data repositories, accessed using bridge
software.

11.2.2 Meta-data History 1990 - 2008
• In the 1990s, tools expanded the scope of the meta-data they
addressed to include business meta-data. Benefits of business meta-
data identified in the industry during this period include:
• The semantic layer between a company’s system “Operational and BI ”.
• Reducing training costs.
• Making strategic information, such as DW, CRM, SCM
• Creating actionable information.
• Limiting incorrect decisions.
• Mid to late 90s, after DW initiatives, and WWW, standardize meta-
data definition and exchange between applications in enterprise:
• CASE (CDIF), EIA, DCMI, ISO 11179, CWM, MDC.
• 21st century saw the update of existing meta-data repositories for
deployment on the web.
• At the current decade proceeds, companies focus on the need and
important of meta-data and expand on how to incorporate meta-
data beyond the traditional structured sources and include
unstructured sources.

11.2.3 Meta-data Strategy
• Statement of direction in meta-data management by the enterprise.
• Acts as a reference framework for development teams “each user
group has its own set of needs from meta-data application”.
• Strategy focus of the key business drivers, issues, and info
requirements for the enterprise meta-data program.
• The objectives of the strategy defined the organization’s future
enterprise meta-data architecture.
• Recommended steps is the logical progression of phase
implementation enable organization to realize the future vision.
1. Meta-data strategy Initiation and planning
2. Conduct Key stakeholder interviews
3. Assess Existing Meta-data Sources and Information Architecture
4. Develop Future Meta-data Architecture
5. Develop Phased MME Implementation Strategy and Plan

11.2.4 Meta-data Management Activities
• As Mentioned before, effective meta-data management depends on
data governance to enable data stewards to set their priorities, guide
program investments, 10 activities:
• 11.2.4.1 Understand Meta-data Requirements
• 11.2.4.2 Define Meta-data Architecture
• 11.2.4.3 Meta-data Standards Types
• 11.2.4.4 Standard Meta-data Metrics
• 11.2.4.5 Implement a Managed Meta-data Environment
• 11.2.4.6 Create and Maintain Meta-data
• 11.2.4.7 Integrate Meta-data
• 11.2.4.8 Manage Meta-data Repositories
• 11.2.4.9 Distribute and Deliver Meta-data
• 11.2.4.10 Query, Report and Analyze Meta-data

Meta-data Activities
11.2.4.1 Understand Meta-data Requirements
• Reflect an understanding of enterprise needs for meta-data.
• Gathered to confirm the need for meta-data environment, to set
scope and priorities, educate and communicate, to guide tool
evaluation and implementation, modeling, standards and services.
• These requirement obtain from both: Business & Technical users in
organization.
• These requirements gathered one by one from:
• Analysis of roles
• Responsibilities
• Challenges
• Information needs of selected individuals in the organization, not from asking
for meta-data requirements.

Meta-data Requirements
11.2.4.1.1 Business User Requirements
• Business users, require:
• Improved understanding of the information from operational and
analytical systems.
• High level of confidence in the information obtained from corporate
DW, Analytical applications, and operational systems.
• Information delivery methods, such as reports, queries, push
(scheduled), ad-hoc, OLAP, dashboards, with high quality
documentation and context.
• The business term “Royalty”:
• Values represent data elements stored in both operational and
analytical systems.
• Appeared in key financial reports, OLAP cubes, and data mining models.
• Business users must understand the intent and purpose of meta-data
management through business requirements “interviews and JAD
sessions”.
• Meta-data management success is the establishment of a data
governance organization.

Meta-data Requirements
11.2.4.1.1 Technical User Requirements
• High-level technical requirement topic include:
• Daily feed throughput: Size and Processing time.
• Existing meta-data.
• Sources – Known and unknown.
• Targets.
• Transformations.
• Architecture flow – logical and physical.
• Non-standard meta-data requirements.
• Technical users include:
• DBAs, Meta-data specialists and Architects, IT support staff, Developers
• They must understand
• the Technical implementation of data “atomic-level details, data
integration points, interfaces, and mappings”
• Business context of the data at a sufficient level to provide the
necessary support.

11.2.4.2 Define the Meta-data Architecture
• Meta-data Architectural layers are:
• Meta-data creation/sourcing, Integration, 1 or more meta-data
repositories, meta-data delivery, meta-data usage, and meta-data
control/ management.
• Meta-data management system must be capable of extracting meta-
data from many sources.
• Meta-data environment should have transparency with single access
point.
• Transparent means that the user can access the data without being
aware of the different environments the data sources “Isolated”.
• Three technical architectural approaches to build meta-data
repository:
• Centralized, Distributed, and Hybrid
• Organizations choose the best suits their needs.

Meta-data Architecture
11.2.4.2.1 Centralized Meta-data Architecture
• Consists of a single meta-data repository contains copies of the live
meta-data from various sources.
• Organization with
• consistency and uniformity within meta-data repository benefit from it.
• limited IT resources or seeking to automate should avoid it.
• Advantages:
• High availability , since it is dependent of the source systems.
• Quick meta-data retrieval, repository and query reside together.
• Resolved DB structure that not affected by the proprietary nature of third
party or commercial systems.
• Improve quality by enhanced meta-data reside in the source system.
• Limitations:
• Complex process may need replication into the repository.
• Maintenance can be substantial
• Extraction could require custom additional modules or middleware
• Validation and maintenance of customized code increase the demands on both
internal IT staff and software vendors.

11.2.4.2.2 Distributed Meta-data Architecture
• Real time approach maintain a single access point into retrieval
engine responds to user requests by retrieving data from source
systems. “no persistent repository”.
• The environment maintains source system catalogs and process user
queries. Object request broker or similar middleware protocol access
these source systems.
• Advantages:
• Meta-data is always as current and valid as possible
• Queries are distributed, improving response/ process time.
• Limited query processing rather than detailed data structures.
• Requiring minimal manual intervention.
• Batch processing is reduced, with no replication or synchronization processes.
• Limitations:
• No enhancement or standardization of meta-data is necessary between
systems.
• Query affected by availability of participating source systems.
• No repository “no ability to support meta-data entries” depend on source
systems.

11.2.4.2.3 Hybrid Meta-data Architecture
• Combined alternative approach, move directly from the source
systems into a repository.
• Benefits from the near-real-time retrieval of meta-data from its
source and enhance it to meet user needs most effectively, when
needed.
• Availability of the source systems is a limit.
• Used to Organization have rapidly changing meta-data, not that with
static and smaller meta-data growth.
• Bi-directional meta-data architecture is approach allows meta-data
to change in any part of the architecture (source, ETL, UI) and then
feed back from the repository into its original source.
• Various challenges with this approach:
• Design forces the meta-data repository to contain the latest version and
force to changes to the source.
• Changes must be trapped systematically, then resolved additional sets
of program / process interfaces to tie the repository back to the meta-
data sources must built and maintained.

11.2.4.3 Meta-data Standards Types
• Two major Types of Meta-data standards:
• Industry or consensus standards
• International Standards
• The international standards are the framework form which the
industry standards are developed and executed.
• Figure 11.2 Show how standards are related and rely on each other
for context and usage.
• The diagram also gives a glimpse into the complexity of meta-data
standards and serves as a starting point for standards discovery and
exploration.

11.2.4.3 Meta-data Standards Types

Meta-data Standards
11.2.4.3.1 Industry / Consensus Meta-data Standards
• The need for sharing common meta-data to support the optimal
usage of shared information has spawned sector-based standards.
• Integration of XML Providing Tools in this standards into a meta-data
management environment still requires custom development.
• Noteworthy industry meta-data standards are:
1. OMG specifications: the creator of CORBA middleware standard and
other related standards such:
• Common Warehouse Meta-data (CWM). Figure 11.3
• Information management Metamodel (IMM)
• MDC Open information Model (OIM)
• XML
• UML
• SQL
• XMI “Extensible Markup Interface”
• Ontology Definition Metamodel (ODM)

Meta-data Standards

Meta-data Standards
• Continuo noteworthy industry meta-data standards:
2. W3C specification: capturing web resources, other associated with URL
3. Dublin Core: Dublin Core Meta-data Initiative (DCMI): focus on
standardization of meta-data for business operational purposes.
4. Distributed Management Task Force (DTMF): set of management and
internet standard technologies developed to mange distributed
computing environment.
5. For Unstructured data: ISO 5964, 2788, 702 , ANSI/NISO Z39.1
6. Geospatial Standards
7. ANZLIC The Australian New Zealand Information Council
8. INSPIRE (Geographic meta-data standards)
9. Industry Sector meta-data Standards have two sample standards :
Automotive industry & Electric Utility industry

Meta-data Standards
11.2.4.3.2 International Meta-data Standards
• The key meta-data standard is ISO / IEC 11179
• The purpose to provide guidance and understanding of the formulation and
registration of data element.
• Relevant parts of the international Standard ISO/IEC 11179 are:
• Part 1: Framework for the Generation and Standardization of Data Elements.
• Part 3: Basic Attributes of Data Elements.
• Part 4: Rules and Guidelines for the Formulation of Data Definitions.
• Part 5: Naming and Identification Principles for Data Elements.
• Part 6: Registration of Data Elements

11.2.4.4 Standards Meta-data Metrics
• Assess user uptake, organizational commitment, and content
coverage and quality.
• Should be primarily quantitative rather than qualitative.
• Some suggested metrics on meta-data environments include:
• Meta-data Repository Completeness
• Meta-data Documentation Quality: for both Automatic and manual
methods.
• Master Data Service Data Compliance
• Steward Representation / Coverage: Organizational commitment to
meta-data.
• Meta-data Usage / Reference: measured of users when using meta-
data repository.
• Meta-data Management Maturity: using the famous model (CMM)
• Meta-data Repository availability: uptime, processing time (batch and
query)

11.2.4.5 Implement a Managed Meta-data Environment
• Incremental steps in order to minimize risks and facilitate acceptance.
• First implementation about managing the meta-data environment is:
• A Pilot: initial implementation cycle that subsequent project from planning,
staff training, Organization changes, and Integration of met-data projects into
IS/IT development methodology is necessary.
• Topics for communication and planning for meta-data management:
• Enterprise Information Management
• Data Governance
• Master Data Management
• Data quality management
• Data architecture
• Content management
• BI / DW
• Enterprise Data Modeling
• Access and distribution of the meta-data

11.2.4.6 Create and Maintain Meta-data
• If a custom solution is developed, creating the data model for the
repository is one of the first design steps after the met-data strategy.
• Creation the repository through: scanning, manual insertion by
authorized users and programs, audit validation, and reports
exceptions.
• Quality of meta-data is critical, low-quality meta-data creates:
• Replicated/ dictionaries/ repositories / meta-data storage
• Inconsistent meta-data
• Competing sources and versions of meta-data “truth”
• Doubt in the reliability of the meta-data solution systems.
• High quality meta-data creates:
• Confident, cross-organizational development
• Consistent understanding of the values of the data resources.
• Meta-data “Knowledge” across the organization.

11.2.4.7 Integrate Meta-data
• Refers to the process begins with acquired meta-data In /out side to
enterprise, extracted, store using tools such as adaptors / scanners, bridge
applications.
• Accomplish repository scanning in two distinct manners.
1. Proprietary interface: single steps scan and load process. “no format-specific
file”
2. Semi-Proprietary interface: Two-step process, scanner produce file,
interface read it.
• Scanning process produces and leverages several types of files during
process:
1. Control files: contains source structure of the data model
2. Reuse file: contains the rules for managing reuse of process loads.
3. Log files: produce during each phase of the process, one for each scan/
extract and one for each load cycle.
4. Temporary and backup files: use during the process or for traceability.
• Use non-persistent meta-data staging area to store temporary and backup
files.
• ETL tools used for DW/BI application are often used effectively in meta-
data integration processes.

11.2.4.8 Manage Meta-data Repositories
• Include a number of control activities to mange meta-data environment,
including:
• Backup, recovery, archive, purging
• Configuration modifications
• Education and training of users and Data stewards
• Job scheduling / monitoring
• Load statistic analysis
• Management metrics generation and analysis
• Performance tuning
• Quality assurance, control
• Quality statistics analysis
• Query/ report generation
• Repository administration
• Security management
• Source mapping / movement
• Training on the control activities and query / reporting
• User interface management
• Versioning

Manage Meta-data Repositories
11.2.4.8.1 Meta-data Repositories
• Meta-data repository refers to the physical tables that met-data
stored using relational DB platform and implementing various
controls and interfaces.
• The content should be generic in design and integrated as possible.
• For example, Business Meta-data “Customer”:
• Could be “Anyone that has purchased a product from our company
within one of our stores or through our catalog”.
• A year later, company adds a new distribution channel “constructs web
site to allow customers order products”
• Business meta-data for customer changes “Anyone that has purchased
a product from our company within one of our stores, through our
mail order catalog or through the web”.

Manage Meta-data Repositories
11.2.4.8.2 Directories, Glossaries and Other Meta-data Stores
• Directory is a type of meta-data store that limits the meta-data to the
location or source of data in the enterprise. “useful for developers and data
super user”.
• Glossary provides guidance for use of terms, and thesaurus can direct the
user through structural choices involve three relationships: equivalence,
hierarchy, and association.
• Multi-source glossary should be capable of the following:
• Storing terms and definitions from many sources
• Representing relationships of sets of terms within any single source
• Establishing a structure flexible enough to accommodate input from varying
sources and relating new terms to existing ones.
• Linking to the full set of meta-data attributes recorded in the met-data
repository.
• Other meta-data stores include
• specialized lists such as source lists or interfaces, code sets, lexicons, spatial
and temporal schema, spatial reference, and distribution of digital geographic
data sets, repositories of repositories, and business rules.

11.2.4.9 Distribute and Deliver Meta-data
• The meta-data Delivery layer is responsible for the delivery of the
meta-data from the repository to the end users and to any
applications or tools required meta-data feeds to them.
• Some delivery mechanisms:
• Meta-data intranet websites for browse, search, query, reporting and
analysis.
• Reports, glossaries, other documents, and websites.
• Data warehouse, data marts, and BI tools.
• Modeling and software development tools.
• Messaging and transactions.
• Applications
• External organization interface solutions (e.g. supply chain solutions).
• Meta-data solution often links to BI Solutions “means of integration
into the delivery of BI to the end user”.
• Finally, meta-data exchanged with external organization through flat
files “XML transportation”

11.2.4.10 Query, Report and Analyze meta-data
• Meta-data guide how “USE ”data assets in:
• Business Intelligence ( reporting and analysis)
• Business decisions ( Operational, tactical, strategic)
• Business semantics ( what we say, what we mean ‘business lingo’)
• Meta-data guide how “MANAGE ”data assets in:
• Data Governance processes to control and govern
• IS implementation and delivery to add, change, delete, and access data.
• Data integration (Operational systems, DW/BI systems)
• Meta-data repository must have a front-end application supporting
the search-retrieval functions to manage data assets.
• The interface provided to business users may have a different set of
functional requirements than that for technical users and developers.
• Some reports facilitate future development such as change impact
analysis, trouble shoot varying definitions for DW/BI projects, such as
data lineage reports.

‏‏‏‏‏‏‏‏Chapter 11: Meta-data Management

More Related Content

What's hot (20)

Similar to ‏‏‏‏‏‏‏‏Chapter 11: Meta-data Management (20)

Recently uploaded (20)

‏‏‏‏‏‏‏‏Chapter 11: Meta-data Management