Data Modeling

The Data Model wizard provides users the ability to quickly and easily build quality data models using raw data from 25+ local and external sources.

IMPORTANT: At least one available tabular instance of Analysis Services (2012, 2014 or 2016) is required in order for the Data Model wizard to run.

Currently, data modeling connectors include:

Files	Local Databases	Cloud and Other	Generic
Excel	MS SQL Server	BI Office (internal queries)	ODBC
Access	IBM DB2	Google Analytics	OLEDB
PowerPivot (Excel)	Oracle	MS Azure SQL	OData
Text	Greenplum	SalesForce
	Hadoop Hive	SharePoint Lists
	MySQL	Twitter
	Netezza	Zendesk
	PostgreSQL	Amazon Redshift
	Teradata	MS Azure Hive
	MongoDb	SAP

Click here for more details on the data connectors.

Overview of Data Model Building

The Data Model wizard facilitates the mash-up of raw data by guiding the user through a step-by-step process in a simple and easy-to-understand way.

For step by step guidance on how to use the Data Model wizard, click here.

Wizard Steps

The main steps are as follows:

• From the Start page, click the yellow "New Data Model" button (click here to see more)
• In the data model wizard panel on the Start page, choose which data sources to add in the data mash-up - you can add multiple files, databases and any combination in between.
• Once the files/databases are imported into the wizard, select one or more specific sets of data to be included from each item. This could be:
  • Worksheets in spreadsheets or named data ranges on a spreadsheet.
  • Tables or views from relational database(s), cloud web data sources etc.
  • CustomQueries against existing data sources (depending on the source type).
• Next select the columns or “attributes” from each data set you want to include in your data model. This can include embellishing the attributes:
  • set attributes as measures
  • split, combine or run calculations against attributes
  • set properties for attributes to effect better analytic capabilities in the main data discovery tool
• Set up any connections or “relationships” between the different data sets. This step allows you to “mash-up” the different data sets and glue them together in your data model.
• Optionally set up hierarchies or 'drill down' trees
• Next, determine the roles and permissions for accessing the data model,
• Last, provide a name and description for the model and process it.

Once the wizard builds the data model you can optionally be launched directly into a new Data Discovery session with the model or you can access it as an existing data source.

Click here for more on the steps and aspects related to modeling.

Built-in Wizardry

Some of the smart logic built into the Data Model engine includes:

• Heuristics - The wizard makes obvious selections and choices for the user to make model building simpler and faster.
• Time Intelligence – With one click, users can add all the common time and date analytic structures to their model based on their raw data
• Relationships – The wizard will automatically reverse engineer any existing content relationships that inherently exist between data sets. This includes working out if certain columns are the “primary keys” for a data set and how they may join up to other data set items.
• Drill-Through – The wizard automatically builds functionality into the model to allow end users to “drill-through” their analytics to the raw, underlying transactional data.

Managing the Data Model

Once the data model is built, you can manage it with the Data Model Manager. Click here for detailed instructions of the Manager.