Creating a layout in Content Analytics Miner

To show analysis results from Content Analytics Miner, you must first create an application layout in Watson Explorer Content Analytics.

About this task

The layout specifies how content in the collection is to be analyzed and how the results are to be rendered, such as in a pie chart or word cloud. You can then publish the layout as an i-Frame widget in Application Builder.

If you need assistance with Content Analytics Miner, use the following steps as a guideline for creating a single-pane layout. Adapt the steps as needed to create a multiple-pane layout, such as a four-pane layout that uses different charts to show correlations, trends, deviations, and connections.

Procedure

To create a single-pane application layout

  1. In the ES_INSTALL_ROOT/default_config/searchserver/repo/analytics/_default/ directory, make a copy of the embedded_template.json file. ES_INSTALL_ROOT is the installation directory, such as /opt/IBM/es/ on Linux or C:\Program Files\IBM\es on Windows.
  2. In Content Analytics Miner, open the application layout customizer by clicking Customize layout in the User menu.
  3. Click Import on the Layout Customizer page, click Browse to locate the copy of the .json file that you created, and then click Import. The layout changes to a single pane layout.
  4. In the navigation pane, click Center Container.
  5. Select the type of widget that you want to add, and then click Add Widget. For best results in your Watson Explorer application, select a widget type that can render all content in one area without scrolling. For example, select one of the RAVE chart widgets, such as Correlation RAVE Chart.
  6. In the navigation pane, select the widget that you added, and then specify conditions for how content is to be displayed. The available choices depend on the type of layout that you create:
    • Select a graphical format for rendering the analysis results, such as a pie chart, bar chart, or word cloud.
    • Select the facet that represents the data that you want to analyze. The facet values are rendered in the selected graphical format.
    • Select the default sort order, such as by correlation value or frequency count.
    • Specify how many of the top N facet values are to be shown in the graphical output.
    • Specify how many of the top N facet values, when sorted by frequency, are to be selected for analysis.
    • If the collection supports contextual views, select the view that is to be analyzed and shown in the results, if any.
    • Specify explicit facet values that you want to analyze only if you do not want the results to show the top N facet values.
  7. Click Apply. Use the Content Analytics Miner to verify that the results of the application layout design match your expectations.
  8. If the design is acceptable, click the Export tab and export the customized .json file.
Parent topic: iFrame Widget Configuration Options