Why Business Intelligence (BI) needs a Semantic Data Model

A semantic data model is a method of organizing and representing corporate data that reflects the meaning and relationships among data items. This method of organizing data helps end users access data autonomously using familiar business terms such as revenue, product, or customer via the BI (business intelligence) and other analytics tools. The use of a semantic model offers a consolidated, unified view of data across the business allowing end-users to obtain valuable insights quickly from large, complex, and diverse data sets.

What is the purpose of semantic data modeling in BI and data virtualization?

A semantic data model sits between a reporting tool and the original database in order to assist end-users with reporting. It is the main entry point for accessing data for most organizations when they are running ad hoc queries or creating reports and dashboards. It facilitates reporting and improvements in various areas, such as:

  • No relationships or joins for end-users to worry about because they’ve already been handled in the semantic data model
  • Data such as invoice data, salesforce data, and inventory data have all been pre-integrated for end-users to consume.
  • Columns have been renamed into user-friendly names such as Invoice Amount as opposed to INVAMT.
  • The model includes powerful time-oriented calculations such as Percentage in sales since last quarter, sales year-to-date, and sales increase year over year.
  • Business logic and calculations are centralized in the semantic data model in order to reduce the risk of incorrect recalculations.
  • Data security can be incorporated. This might include exposing certain measurements to only authorized end-users and/or standard row-level security.

A well-designed semantic data model with agile tooling allows end-users to learn and understand how altering their queries results in different outcomes. It also gives them independence from IT while having confidence that their results are correct.

Data Modeling – Fact Table Effective Practices

Database Table
Database Table

Here are a few guidelines for modeling and designing fact tables.

Fact Table Effective Practices

  • The table naming convention should identify it as a fact table. For example:
    • Suffix Pattern:
      • <<TableName>>_Fact
      • <<TableName>>_F
    • Prefix Pattern:
      • FACT_<TableName>>
      • F_<TableName>>
    • Must contain a temporal dimension surrogate key (e.g. date dimension)
    • Measures should be nullable – this has an impact on aggregate functions (SUM, COUNT, MIN, MAX, and AVG, etc.)
    • Dimension Surrogate keys (srky) should have a foreign key (FK) constraint
    • Do not place the dimension processing in the fact jobs

Related References

Data Modeling – Dimension Table Effective Practices

I’ve had these notes laying around for a while, so, I thought I consolidate them here.   So, here are few guidelines to ensure the quality of your dimension table structures.

Dimension Table Effective Practices

  • The table naming convention should identify it as a dimension table. For example:
    • Suffix Pattern:
      • <<TableName>>_Dim
      • <<TableName>>_D
    • Prefix Pattern:
      • Dim_<TableName>>
      • D_<TableName>>
  • Have Primary Key (PK) assigned on table surrogate Key
  • Audit fields – Type 1 dimensions should:
    • Have a Created Date timestamp – When the record was initially created
    • have a Last Update Timestamp – When was the record last updated
  • Job Flow: Do not place the dimension processing in the fact jobs.
  • Every Dimension should have a Zero (0), Unknown, row
  • Fields should be ‘NOT NULL’ replacing nulls with a zero (0) numeric and integer type fields or space ( ‘ ‘ ) for Character type files.
  • Keep dimension processing outside of the fact jobs

Related References

Data Modeling – Database Table Field Ordering Effective Practices

Field order can help the performance on inserts and updates and, also, keeps developer and users from having to search entire table structure to be sure they have all the keys, etc.

Table Field Ordering

  1. Distribution Field Or Fields, if no distribution field is set the first field will be used by default.
  2. Primary Key Columns (including Parent and Child key fields)
  3. Foreign Key Columns (Not Null)
  4. Not Null Columns
  5. Nullable Columns
  6. Created Date Timestamp
  7. Modified (or Last Updated) Date Timestamp
  8. Large text Fields
  9. Large binary Columns or Binary Field references

Related References

Where do data models fit in the Software Development Life Cycle (SDLC) Process?

In the classic Software Development Life Cycle (SDLC) process, Data Models are typically initiated, by model type, at key process steps and are maintained as data model detail is added and refinement occurs.

The Concept Data Model (CDM) is, usually, created in the Planning phase.   However,  creation the Concept Data Model can slide forwarded or backward, somewhat, within the System Concept Development, Planning, and Requirements Analysis phases, depending upon whether the application being modeled is a custom development effort or a modification of a Commercial-Off-The-Shelf (COTS) application.  The CDM is maintained, as necessary, through the remainder of the SDLC process.

The Logical Data Model (LDM) is created in the Requirement Analysis phase and is a refinement of the information entities of the Concept Data Model. The LDM is maintained, as necessary, through the remainder of the SDLC process.

The Physical Data Model (PDM) is created in the Design phase to facilitate the creation of accurate detail technical designs and actual database creation. The PDM is maintained, as necessary, through the remainder of the SDLC process.

Related References: