Assign Semantics And Synthesis Strategy- Process
As an example, if we choose a table from the list of available fields in the left pane and select "Personal Details," the system will search for all personal identification information (PII) associated with that table. It will then display the columns in the table that suggest synthesis strategies.
All columns associated with the table, such as name, date of birth, email address, gender, phone number, etc., will be displayed under "Personal Details."
When clicking on the "Name" column, the following options will be available to fill out:
Tokenize
Assign Semantics
Pattern
Dependent Field
The Assign Semantics field provides a list of options in a dropdown menu (e.g. Date, Address, Phone Number, etc.) that can be selected to assign meaning or context to the data values in the selected table columns.
If the "Name" column is chosen from the table, then the "Name" option should only be selected in the "Assign Semantics" dropdown menu. Choosing any option other than "Name" in the "Assign Semantics" dropdown menu may result in incorrect data synthesis.
After selecting the "Name" option in the "Assign Semantics" dropdown, a pattern needs to be chosen for synthesizing the data. The "Pattern" dropdown includes options such as "Person Name," "First Name," "Last Name," "Random," "By Religion," "By Nationality," and so on.
Based on the pattern selected, the data will be synthesized accordingly. For instance, if the "Person Name" pattern is selected, the data synthesized will be a random person's name. Similarly, if the "By Religion" pattern is selected, the synthesized data will be based on common names associated with a specific religion.
If the desired pattern is not available in the dropdown, there is an option to add a new pattern.
When clicking on "Add New Pattern," a popup window will appear, where the user can see the semantic group name and select a strategy for the new pattern.
After selecting the pattern and filling out the necessary fields, the "Test" button can be used to verify the semantic types assigned to the data attributes and preview a sample of the generated data. Upon clicking the "Test" button, the system will generate a preview of the synthetic data based on the assigned semantic types and constraints. The preview will then be displayed on the screen for the user to review before and synthesized data.
After testing the synthesis, the "Save" button can be used to save the job.
The "Clear" button can be used to discard any changes made to the job and redirect the user back to the jobs page.
After saving the data, the user can click on the "Next" button to proceed to the next page and confirm the synthesis strategy.
Last updated