Generating Page Insights in Publish Pro

A "Page insight" is a natural language analysis of a page that is dynamically generated at build time using AI. Insights can be added to your page as variable text fields, and provide your reader with a textual analysis of the content of the current page. They also reflect the filtered data in the page; for example, if you are building a PDF that describes the Sales in a particular region, the filter applied to select that region at build time will also affect the insight information where the page contains filtered visuals.

You can add a variable text field that includes a #PageInsight tag to any page in your publication. Page insights are generated in your Application Language and contain analytic information about the content (visuals and applied slicers) in the built page.

It is important to note that the content is generated each time Pyramid builds your publication, and may be different every time.

Note: You can only make use of this functionality if your administrator has selected the Enable Generative AI checkbox in the AI Settings page of the Admin Console. If your administrator does not select this option, the function is not available for use.

Warning: When using LLMs, your assets are generated using public domain algorithms. This can produce erroneous and inconsistent or random results. Use at your own risk.

Page Insights Content

When you open a page in your output publication, any Page Insight variable text field is replaced with a block of generated text in your Application Language. The text is "analytic," describing the content of the page that the variable appears on including applied filters. It is typically a paragraph or two, although this may depend on your Settings.

What does the insight look like?

The #pageInsights tag is replaced with a paragraph of analytic text when the publication is built.

ClosedExample: Page Insights in a PDF

The replacement text describes visuals on the page and reflects any filtering that was applied at build time:

What content does the insight contain?

The insights are generated in the current user's chosen Application Language, including being presented as right-to-left text where that language is Hebrew or Arabic. The analysis content includes information from the following categories:

  • Basic Facts: Standard metric analysis, including the average, maximum, minimum, dispersion, median, and standard deviation.
  • Pareto Analysis: The value's total, top n, and bottom n that account for a certain percentage.
  • Relationships: Any positive or negative metric correlations, relationships, and a correlation statistic.
  • Trends: Uses regression analysis with r-squared and slope coefficients to explain how changes to a metric affects another metric, the closeness of fit score, and the slope correlation
  • Forecasts: Best fit forecasting (across eight different algorithms) if the query contains a date-time hierarchy to predict whether a value will increase or decrease over a certain period and by how much, with the range of fluctuation
  • Outliers: Displays the top three most outstanding data points in the query. It also shows the probability density for each outlier.
  • Categorical Analysis: Depending on the amount of data in the query, categorical insights may display information about comparatives, correlations, and forecasts.

Fairness analysis

If you are using an LLM to generate Pyramid content, and your administrator has selected to enable Fairness Analysis in your AI settings, then some fairness analysis content will be generated by the LLM and added to the content in your Smart Insights panel or any Slide or Page Insights added to your presentation or publication, respectively.

The fairness content, generated by the LLM, is created through an analysis of the measures and attributes used in the discovery (but not the values of those measures and attributes) and provides insights about any bias that may disproportionally harm or benefit any particular group, particularly historically disadvantaged groups. The returned content is generated by the underlying LLM.

Important: The fairness content will be regenerated by the underlying LLM whenever the Smart Insights panel is opened, at runtime whenever the presentation is refreshed, or whenever a new publication is built. Each time it is regenerated, the content may change.

Adding Page Insights

Open the page

Either open the page that you want to add your insights to from the Pages panel or create a new page from that same panel.

Add insights manually

From the Toolbox:

  1. Right-click Add Text Field > Page Insights.

  2. Click the canvas at the point where you would like to add the text field.

    3. Clicking creates a standard sized text field. Alternatively, you can drag your cursor across the canvas to draw a box indicating the size of the item when dropped.

  3. Select the #PageInsights text field on the canvas.

    4. The Page Insights panel opens at the bottom of the page.

  4. Optionally, specify the settings for your Page Insights.

    5. These options allow you to control the tone type and verbosity used by the LLM when generating the replacement text. For more information, see Settings.

  5. Click Apply.

Tip: You should remember that the generated insights information will use up a lot more space than the variable it is replacing. You might, therefore, want to make the text box quite large and manually build a PDF to check the amount of text you might need to accommodate.

Add insights using the Chatbot

If you include a request to add Page Insights explicitly to your prompt (you say or type "add insights" into the Chatbot), a Page Insights text field (a text field containing the variable #PageInsights) is added to your publication as part of the process.

  • Click here for more information about the Chatbot
  • Click here to learn about creating a publication using the Chatbot

Editing and Deleting #PageInsights

You can edit and delete #PageInsights as you would any other text field or variable text field. Be aware that the #PageInsights tag will be replaced with a considerable amount of text, and plan its appearance and position in your publication accordingly.

  • Click here for more information

Settings

Open the Settings dialog by clicking the variable on the canvas:

These settings are used by the LLM when generating the Insights:

  • User Tone Type: Sets the appropriate tone for the insight message. This can be one of a set of tone values ranging from Professional to Whimsical.
  • LLM Verbosity: Sets the level of detail to use for the explanation in the generated insights. This can be one of Low, Medium, or High.

Tip: If you *always* want to select a particular Tone type or Verbosity, you should reset your user defaults to save time in future. For more information, see User Defaults.