Cluster documents in designed pages - Fluid Topics - Latest

Fluid Topics Designer Guide

Category
Reference Guides
Audience
public
Version
Latest

To cluster documents in the Search results component, proceed as follows:

  1. Go to Administration, and select Design, then Search results.
  2. Open the Default search.
  3. Select or add the Search results component.
  4. Select the Settings tab (cogwheel) of the Search results component.

  5. In the Search results section, select a metadata key and its corresponding Result template.

    • The Search results component can show individual results differently based on their metadata. For example, it is possible to create a template for tutorials, and another for reference guides.

    If no template is selected, the default template applies.

  6. Go back to the Search results menu, and select a Result template.

  7. Add one or multiple Result metadata components.
  8. For each Result metadata component, select a metadata key.

    Users can choose a document variant based on the metadata key. For example, for a metadata key of audience, users can choose between public, internal, etc.

  9. For each metadata used for clustering, select the Use for clustering results check box.

    It is possible to choose if a document opens automatically after selecting a clustering metadata by selecting the Go directly to document when selecting a cluster check box.

When using the designed Search results page, it is now possible to select a specific version of a document by selecting the desired metadata value.

A search result titled 'Fluid Topics Configuration and Administration Guide' with a date '2024-02-01' on the top right. Underneath, there are three buttons. The first one for the audience says 'public' with a dropdown showing options for 'public', 'internal', 'onprem', and 'premium'. The second one says '4.3' referring to the version of the document. The third one says 'markdown', referring to the format of the document.

In the case of complex clusters with multiple clustering metadata, some results might not be directly accessible.