Step 1: Classification Name and Sources
After clicking the Auto-Classify link in a page of Watson™ Explorer Engine search results, a screen like the one shown in Figure 1 displays.
The Classification Name and Sources section is the only active portion of the Auto-classification screen at the moment. Enter or modify the Classification Name, which is the name of the auto-classification hierarchy that you want to create. The Watson Explorer Engine enables you to create and save multiple auto-classification hierarchies, which are also referred to as classifications.
The current classification, listed at the top of the screen below the Auto-Classification header, reflects the current or suggested name for the classification that you are creating or modifying. The default name for your first classification run is the same as any Display name that you set in the Meta Information for the source for that collection, or the name of your search collection if no display name was set. In this case, the search collection used by this tutorial has the display name Enron Email (which we assigned in Customizing the Source for Your Search Collection). You can optionally change the Classification Name, or specify a new one after auto-classifying your data, enabling you to create multiple classifications for the sources used by one application.
The Sources area is automatically populated with sources from your project if you came from a results page. If you navigated to the auto-classification tool directly, there will be a dropdown containing all sources in your installation. If more than one source is involved in your project, make sure that all of the sources that you want to auto-classify are checked.
The Project area should be automatically filled with the project auto-class-tutorial.
After making any changes that you want to make, click Save to save the name of your classification and the sources that it will organize into classes. The dialog shown in Figure 2 displays.
Selecting the Return to the configuration tab option takes you to the next step of the automatic classification process, which is discussed in Step 2: Create Classes. Selecting the Go to the sample results tab link takes you to a sample result screen on which you can create classes and sub-classes manually.
Manually created classes and sub-classes can be used to either entirely define the class set, or to refine the classes produced by using a taxonomy or auto-sampling. When manually creating classes, it is important to have an understanding of the collection that is being classified. Each query can be run to identify the number of results, but without some knowledge of the collection, identifying the correct queries can be time consuming.
Editing Classes Manually explains how to manually create and manipulate classes.
For this tutorial, we will create classes and sub-classes automatically.
To proceed to the next section, click Step 2: Create Classes.