About Managing Data Model Indexing, Synonyms, and AI Descriptions Using a CSV File

Administrators can manage subject area indexing, synonyms, and AI Descriptions for Oracle Analytics AI Assistants and home page searches using the Search Index page. Alternatively, administrators can export the data model configuration to a CSV file, update it, and import the file into Oracle Analytics.

Administrators can use the CSV file to:

  • Migrate AI-ready subject area metadata between environments.

  • Work with subject-matter experts to add or update synonyms and AI Descriptions for assigned data model columns.

  • Update metadata for multiple columns more efficiently than editing each column individually in the Console.

Administrators can share the complete CSV file or separate files that contain only the columns assigned to each subject-matter expert. After the contributors update the files, the administrator imports the completed files into Oracle Analytics.

When you open the exported CSV file in an editor such as Microsoft Excel, update only the supported indexing, synonym, and AI Description columns. All changes in the CSV file are imported into Oracle Analytics. Don’t change values in other columns, because doing so might introduce errors into the data model or break visualizations.

About Specifying Indexing Properties and Synonyms

You can specify the following indexing properties and semantic metadata in the CSV file.

  • EnableAssistant - This cell is set to Y if the Assistants checkbox in selected in the Data Model tab of the Search Index page. The cell is blank if the object or column's Assistants checkbox isn't selected.

    In the CSV file, set this column to Y to make the object or column available to Oracle Analytics AI Assistants, or leave it blank to exclude it.

    When you set this column to Y for an object, you must also set it to Y for any child elements that you want the Assistant to use.

    When you set this column to Y, you can specify an IndexType, add synonyms, and add an AI Description.

  • IndexType - The IndexType cell specifies how Oracle Analytics indexes the object or column. The value corresponds to the indexing option selected on the Data Model tab of the Search Index page.

    You can specify these values:
    • Index Metadata Only - Indexes only dimension, attribute, and measure names. In the Console this option is displayed as Name. You must set the EnableAssistant cell to Y to specify this option. Use this option for columns that contain sensitive values that you don’t want users to see in their home page search results or Assistant responses.
    • Index - Indexes metadata and data values. This option applies only to dimension or attribute columns. In the Console this option is displayed as Names and Values. You must set the EnableAssistant cell to Y to specify this option.
    • No value - Excludes subject areas, tables, or columns from the index.
  • DataSecurity - This cell is set to Y when the Sensitive Data option is selected for the column in the Data Model tab of the Search Index page. Set this column to Y to restrict indexed data to users with access to that column.

  • EnableHomePageAsk - Indicates whether the home page search can use the subject area data. This cell is set to Y or left blank depending on the setting for the data model root on the Data Model tab. If the data model root is selected for the home page, EnableHomePageAsk is set to Y for all child rows. If the data model root level is not selected in the Data Model tab, this cell is blank for all rows.

  • Synonyms - The Synonyms cell contains alternate names for a data model column. Confirm that the corresponding EnableAssistant cell is set to Y and enter synonyms, using a comma to separate multiple synonyms. The cell contains any synonyms defined in the Synonyms field in the Data Model tab (Search Index page). See About Adding Synonyms to Your Data Models and tips-using-synonyms-search.html#GUID-EC4F4D27-9397-4FBA-8A6B-170FF9689E01.

  • AI Description - The AI Description cell contains business context that helps Oracle Analytics AI Assistant interpret a data model column correctly.

    Use an AI Description to explain information that the column name and synonyms don’t convey, such as a measure’s unit or scale, a value format, an aggregation rule, or a relationship with another column. For example, an AI Description might explain that Fiscal Quarter uses a different format depending on the fiscal calendar, or that Invoice Type applies only to invoiced amounts, not forecasts.

    Write concise, specific descriptions that explain the business information needed to interpret the column correctly. Don’t repeat the column name or use the AI Description only to list alternate names. Use the Synonyms column for alternate names. For more information, see About AI Descriptions for Subject Areas.