Test sets

A test set is a document set that is added manually for the purpose of testing your separation, classification, or extraction settings. The documents in your test set are typically real examples of the type of documents you want to process with this project. As a result, testing your configurations is a good indicator of whether more work is required.

You can also use a test set when configuring your extraction settings. Testing locators with test documents in the Document Viewer enables you to see the expected extraction results as you change the configuration.

You can also compare the results of your test documents against a benchmark set. This shows how your project configuration performs on a set of test documents compared to a set of golden files. Any significant differences in the results may indicate that the current settings may need additional tweaking.

Note For the best results, ensure that the documents used in any document set are located on an NTFS file system. A FAT 32 file system has a restriction on the number of files allowed, and exceeding this number could result in the loss of documents and data. For more information, see http://technet.microsoft.com/en-us/library/cc776720%28v=WS.10%29.aspx.

You can manage your test document sets in the following ways: