Associative Search Databases

This uses a fuzzy search algorithm for database searches. It is able to better differentiate between the best alternative and the second best alternative than a fuzzy database. As a result, there are fewer false negatives than you might get with a fuzzy database. However, there are some limitations to this type of database.

Tip When using this type of database as input for the Database Locator, it can be beneficial to amend the Regions settings. By default, all pages are taken into account and the entire text of every page is searched. These settings will work adequately when using an Associative Search Database for any document type.

However, if you are processing invoices, the default settings for regions are not ideal. It is recommended to define a region that searches the top 35% of a page and a second region that searches the bottom 30% of each page in a document. Searching within these regions only can provide a better overall accuracy and performance for extraction when processing invoices.

Apply these settings and thoroughly test your extraction results. Modify the region sizes until you see the best results.

To manage associative search databases: