Categories & Doc Types Overview
Overview
Edith is trained to organize asset documents into 17 categories across 3 Asset types (Wells, Facilities, and Pipelines). These categories are available across all asset types which ensures that all documents can be properly categorized.
Please refer to the following sections to see how Edith has been trained to categorize specific document types and by file extension.
Note: This may not include every document type that is currently in Stack nor is an exhaustive list of every document type that could exist. Our team reviews new documents periodically to identify new document types for Edith to be trained on.
If your uploaded document isn't categorized as expected or has an incorrect document type applied, please contact us at [email protected] for assistance.
Edith Tips & Tricks!
Individual Documents: Edith performs best with individual documents. Edith learns from the words on the first page, so it is best practice to avoid scanning multiple documents together. If you do scan documents together, the category will typically be determined by the first document in the scan.
Cover Letters: To keep a cover letter with your document but prevent it from being categorized as correspondence, attach it as the last page.
Accuracy: Edith's accuracy improves with more data. Common form-type documents are excellent training sets, but not all documents fit this mold. You can manually update categories, document types, titles, or dates to help Edith learn for more uncommon doc types we find related to assets.
User Feedback: Our team reviews user changes to categories, document types, titles, and dates to identify areas where Edith may need more training. If you see multiple instances of a document type with incorrect categorization, reach out to [email protected] for assistance.
Last updated