Creating a Semantic Type

To add a new semantic type to the Business Glossary, perform these steps:

  1. On the Spectrum Discovery menu bar, click Define Glossary > Semantic Type.
  2. Click the Add Semantic Type icon .
  3. On the Create Semantic Type page that is displayed, specify these details.
    Fields Description
    Name Enter the name of the semantic type. Example: Address.
    Note: Entering this detail is mandatory.
    Description Describe the semantic type.
    Tags Specify a tag for the semantic type. You can select one from the existing list or create a new tag. Tags are useful in categorizing the semantic types. You can also enter multiple tags for a semantic type.
    Business Rule(s) Describe the rules associated with this semantic type. In case of multiple rules, use line separation.

    Example: Select regex patterns for any pattern that needs to be there
    Data Type Specify the data type of the semantic type.
    Note: This is a mandatory detail.
    Rule type The rule type you define here will be used during profiling to find matching records. Select one of these options, as required.
    1. None: Select this if you do not want to use any rule type association for this semantic type.
    2. Matches: Specify the regular expression based on which the records will be fetched.
    3. Exists: Specify the data to be searched for in the entire data set. For example, if you specify Single, Married here, when this rule is used on the Profiling page, it will look for it in the data set, fetch all the records that have these values and present it on the Data Profiling Results page. For more details on semantic analysis see the section Semantic Analysis.
      • You can enter the required data in the text box.
      • You can also import data using a .txt or .csv files. Click the import icon for it.
        Note: Ensure the file contains single column and does not have a header.
        1. In case of multiple columns, only comma delimited values will be supported. For such files, only the first column will be imported. For any other delimiter, the first 100 rows will be uploaded.
        2. If your file contains more than 100 records, only the first 100 records will be imported.
  4. Click Save.
    The semantic type is saved and displayed on the Semantic Type tab of the Define Glossary page.