InfoSphere® Data Architect helps data modelers to design, relate, and standardize
data models.
Key features for designing data models
InfoSphere
Data Architect supports data modeling for several data server types.
You can perform the following essential data modeling tasks:
- Logical and physical data modeling
- Create logical data models that describe how your organization
wants to collect data and how data objects are related to each other.
Then, transform to a new physical data model, or reverse-engineer
from an existing source. Use the physical data model to model how
data should be stored.
InfoSphere Data Architect also supports dimensional
modeling for both logical and physical data models:
- Create dimensional models to model multidimensional data. Dimensional
modeling supports facts, dimensions, hierarchies, and outriggers.
- Easily identify dimensional objects in the Data Project
Explorer. For example, a fact table, SLS_SALES_FACT appears in the Data Project Explorer as SLS_SALES_FACT [FACT] in a dimensional data model.
- Enable dimensional notation for a logical or physical data model,
then discover the facts and dimensions in the model to create dimensional
data models.
- Data model transformation
- Transform between data model types:
- Transform a logical data model into a new or existing physical
data model.
- Transform data models into other model types: UML, Cubing Services
or Cognos®, logical-dimensional,
physical-dimensional, and XML schema.
- Create new data models or reverse-engineer from a source
- Create new physical data models from scratch, or generate a physical
data model from an existing data source. Use the physical data model
to design or build upon the physical structure of your data objects
and their relationships.
You can also reverse-engineer from an existing IBM® Industry Model into a Netezza-based
physical data model.
- Import and export data models
- Import models from a supported tool, or export to a supported
tool.
- Visualize your data models
- Create diagrams for logical, physical, and dimensional data models
to help yourself and others understand data objects and how those
objects are related.
- Deploy changes to a target database
- Generate DDL to deploy your physical data models and updates to
physical data models to a data server, then run the DDL. Users, data
modelers, application developers, and other business process owners
can then work with the new and changed objects on the data server.
- Compare and synchronize changes
- Compare a data model or data object to another data model or data
object. Compare a data model or data object to the original source
data model or object. Compare data sources to one another. Easily
identify and merge differences.
Key features for relating data assets
You
can use InfoSphere Data
Architect to perform the following data integration tasks:
- Discover relationships between databases (source and target schemas
or related tables)
- You can use the enterprise metadata tools that are provided with InfoSphere Data Architect
to find and map relationships among a variety of data sources, build
scripts that represent those relationships, and then deploy scripts
to federated, non-federated, local, or remote servers.
- Map relationships between data models
- InfoSphere Data Architect
provides a mapping editor that you can use to visually define relationships
between elements of one or more data sources and a single target in
a data design project. The data source can be any physical data model
or logical data model in the project. The target can be a physical
data model, a logical data model, or an XML schema (XSD file).
- Build expressions that provide integrated views across data objects
- Define criteria, such as functions, join conditions, filter conditions,
and sort conditions, that you can add to the mapping model. You can
then use the expressions in scripts that you deploy.
- Generate scripts from mapping models to integrate sources and
targets
- Generate a script from the mapping model that you can deploy.
You can use the script to transform and filter data from mapping-model-compliant
sources to mapping-model-compliant targets.
- Access remote sources that are defined to a federated database
- You can access remote data sources that are defined to a federated
database. InfoSphere Data
Architect supports connecting to a WebSphere Information Integrator
federated database and accessing the remote data sources that are
defined to the federated database.
InfoSphere Data Architect supports JDBC
access to the following data sources:
- DB2® family
- Informix®
- Microsoft SQL Server
- Oracle
- Sybase
- Teradata
InfoSphere Data Architect also supports ODBC data sources, including
Excel and WebSphere II Classic Event Publisher (by using an ODBC connection).
- View peer relationships between objects
- The Topology view in InfoSphere Data Architect displays a read-only graphical view of the peer relationships
between objects in your enterprise. This view shows peer relationships
between databases, schemas, and tables. For example, you can see all
of the objects in an InfoSphere Information Server federated server and the federated databases
that it connects to. You can print the contents of the Topology view
to share this information with other team members.
Key features for standardizing data assets
You can use InfoSphere Data Architect’s data standardization features to improve data quality
and enhance understandability. These features are especially critical
for large organizations with multiple team members.
InfoSphere Data Architect
has three main features that allow teams to enforce organizational
standards:
- Domain modeling
- A domain model consists of a group of domain data type objects.
A domain object represents an abstract data type that can be restricted
by adding constraints. Domain data types are based on base data types.
For example, you can define domain data types for commonly used definitions
such as social security number, sex, height, or marital status.
You can use InfoSphere Data Architect to create domain models and atomic domains, and then
share these models among team members who are working on logical or
physical data models.
InfoSphere Data Architect provides tooling that helps you create
and manage domain models. You can create a domain model from a wizard,
either from an existing template or from scratch.
- Glossary and naming models
- When a naming standard is enforced, terms defined in a glossary
must be used and term order must be followed when you are naming a
data object. InfoSphere Data Architect provides tools that help you
define glossary and data object naming standards, to make it easy
to create data object names that are compliant with these naming standards.
You can also use a glossary model to validate existing data objects,
either in a model or on a database, to ensure their compliance with
naming standards.
You can use a glossary model in the following
ways:
- During the analysis of a data model, to determine whether the
model adheres to naming standards.
- During data model object modification in the Properties view,
to provide content assistance when specifying object names.
- During logical to physical data model transformation, to determine
how to name physical data objects.
- During physical to logical data model transformation, to determine
business terms based on a physical data object.
- During the creation of a mapping model, to allow the discover
function to more easily relate source and target elements.
You can use InfoSphere Data Architect to generate glossary models, and then share these
models among team members who are working on logical or physical data
models.
- Data model analysis
- At any time when you are building a data model, you can analyze
the model to verify that it is compliant with the defined constraints.
You can also analyze existing databases. Before you analyze a data
model, you can set preferences for data model analysis. Then you can
use a wizard to specify additional model analysis settings and perform
the analysis. Results are displayed in the Problems view, from which
you can navigate to listed objects that are not in compliance.
You
can analyze a model that is stored in a data project, or you can analyze
a database that is displayed in the Data Source Explorer.