Smart Naming Conventions

ENABLING SMART NAMING CONVENTIONS

The Smart Naming Conventions module is sold separately, please contact your Account Manager to discuss activation.

USE CASES FOR SMART NAMING CONVENTIONS

Within large and diverse datasets, it's very easy for minor errors to creep into what should be a consistent pattern, and equally easy for these errors to compromise or even invalidate data extracts for analysis. These minor discrepancies can be hard and time-consuming to detect manually. Smart Naming Conventions will both check for, and where possible correct, compliance with a set of rules to ensure data consistency:

• Automate the division of a single string Dimension containing multiple data points into its constituent elements (eg. "REGION_CAMPAIGN_ID_UNITS" into "REGION", "CAMPAIGN", "ID", "UNITS").
• Identify and flag missing data:
  
• Correct misaligned datasets:
  

START THE SMART NAMING CONVENTIONS WIZARD

1. In the Connect Element, navigate to any Datastream's Overview page.
   
2. You will see the Smart Naming Conventions tile on the left, below Scheduling.
3. Click +Add naming convention to launch the wizard that will guide you through the rest of the process.

CREATE SMART NAMING CONVENTIONS

1. Select Column
   You will be asked to select the Datastream column to which you want to apply a naming convention. This column must be formatted as a string, and mapped as a Dimension.
   
   
2. Exclude
   The platform will generate a preview of the values within the selected column. This sample will be analyzed by the system in order to establish a naming convention, and here is your opportunity to excise any column entries that do not comply with your chosen convention. Un-tick the checkbox next to any Dimension you want to remove from consideration when creating the convention.
   Please Note: In order to create accurate rules, it is very important that the sample doesn't include any inconsistent values, or the rules created won't be reliable.
   
   
3. Rules and Delimiters
   • Delimiters Detected
     Delimiters are the characters that the system uses to identify breaks between fields. Here you will see the characters identified by the platform and as with Rules you can disable individual delimiter characters, and the page will auto refresh to show the results of that change in the Split Preview (see below).
     Please Note: The system requires at least one delimiter to be active, and can have no more than 3 delimiters in a single column.
     For more information on supported delimiter characters and date formats, see section Supported delimiter characters and date formats.
   • Rules Detected
     You will see the list of logic statements that the system has compiled to govern the process of creating and enforcing the column field separation. If the convention is not quite right, you can disable individual rules, and the page will auto refresh to show the results of that change in the Split Preview (see below).
     To edit the list of permitted values, click Edit. The maximum number of elements in the list is 1000. The values are case-sensitive.
   • Split Preview
     This table shows a preview of your data extract as it will appear post-application of the new naming conventions and field divisions.
     
     
4. Set Headers
   The next stage is to set a header title for each of your new columns. Each will be presented along with an example field value as an aid to memory. Each new column can be named exactly as you would any other column header in your dataset.
   
   
5. Receive Errors
   If the selected column contains a value which cannot be interpreted in any way through the convention's rules, the fetch process will stop and a notification will be flagged in the Datastream Overview.
   • Send Emails
     Tick this checkbox to also receive a copy notification to the email address specified in your Datastream Settings.
   • Continue Processing
     Tick this checkbox and fetches will not be stopped when an error occurs, although the relevant notifications will still be triggered.
     
     
6. Receive Warnings
   When the dataset contains information that is not wholly consistent with the entire naming convention, but can still basically be parsed, Adverity will attempt to rectify minor errors and align the dataset with accepted values (e.g. skipping a missing column value to maintain other column values). 
   Here, you can individually select which columns you wish to trigger notifications when this occurs by switching the sliders on or off.
   Please note: If both the below notifications are deactivated, the system will still correct inconsistent data, but you will receive no notifications at all that it has taken place.
   
   • Send Emails
     Tick this checkbox to also receive a copy notification to the email address specified in your Datastream Settings.
   • Raise Warnings
     Tick this checkbox and a Warning notification will appear in the Issues section of the Datastream overview for any columns with the sliders turned on.
     
     
7. Summary
   Once you've finalized your settings, you'll be shown a brief recap of the rules you have set for this convention. If any are missing or incorrect, you can skip back in the process to rectify them,
   Click Create+ and your new Naming Convention will be saved and activated on the Datastream. The Smart Naming Convention tile will show an "active" tag, and the convention rules themselves can be found within the Transformation Scripts tile.
   Note: Clicking on the SNC entry in the Transformation Scripts tile will show you the summary of your configuration settings.

Reference

Supported delimiter characters and date formats

The table below summarizes the list of supported delimiter characters.

The following date formats are supported:

• %d/%m/%Y
• %d-%m-%Y
• %d_%m_%Y
• %d.%m.%Y
• %m/%d/%Y
• %m-%d-%Y
• %m_%d_%Y
• %m.%d.%Y
• %Y/%m/%d
• %Y-%m-%d
• %Y_%m_%d
• %Y.%m.%d

FAQS

I have a naming convention, but I need to add another value to the list of accepted options, how do I add it?

Smart Naming Conventions can only be created by reviewing historical data, and as such can not be adapted ahead-of-time for new incoming field values. In order to account for new values:

1. Remove the old convention's Transformation script.
2. Refetch a dataset that includes the new value.
3. Run the Smart Naming Conventions wizard again. This will create a new set of rules that include the new values.