Lexicon: Frequently Asked Questions¶
Alation Cloud Service Applies to Alation Cloud Service instances of Alation
Customer Managed Applies to customer-managed instances of Alation
Lexicon Job Behavior¶
No. The Lexicon job processes metadata for all data objects in the catalog. It cannot be scoped to a specific data source or schema.
Lexicon runs automatically every Sunday at 8:00 AM, or you can run it manually on demand.
No. Changes made to suggested expansions on the Lexicon page or on catalog pages will only be applied after the next Lexicon job run.
Auto-Titling and User Input¶
No. Titles must be confirmed or discarded one by one on the catalog or Lexicon pages.
No. Titles manually added or confirmed by a user will always take precedence. Auto-titles are only suggested for blank title fields.
ALLIE generates title suggestions based on detected abbreviations in data object names, following these rules:
It will not override user-entered or confirmed titles.
It will only suggest a title if it has sufficient confidence in all parts of the name.
If you reject an expansion for an abbreviation in Lexicon, it will no longer be used in title generation.
There are a few possible reasons:
The Lexicon job has not been run yet.
Alation lacks confidence in making a guess due to:
Use of non-alphabetic characters (for example, numbers and symbols)
Multiple expansions for a single abbreviation, leading to ambiguity
Uncertain parts within a name (for example, conflicting or vague abbreviations)
Yes. Confirmations or rejections made on catalog pages are included in Lexicon feedback and affect future title suggestions after the Lexicon job runs.
You can improve suggestions by:
Adding new abbreviation-expansion mappings
Confirming or editing existing mappings
These actions can be performed on the Lexicon page by Catalog and Server Admins.
Changes take effect only after the next Lexicon job run.
Lexicon Content and Structure¶
No. Currently, the Lexicon only displays all expansions for each specific abbreviation.
Nothing. The Lexicon is built from data object names discovered during metadata extraction (MDE). You must connect and extract metadata from data sources to populate it.