CMS 10.6 Asset Text Search Configuration
Prerequisites:
CMS 10.6 must be installed. See CMS Statistics to check the site version.
Users must have system administrator permissions.
CMS 10.6 provides asset text search functionality. After configuration, users can access the utility bar search feature to run text queries on text-based assets (DITA, XML, HTML, markdown, etc.). Users can also access this feature via the Oxygen Desktop Plugin.
Asset text searches are disabled by default. For users to run asset text searches, the CMS requires system administrators to configure the feature.
This section includes:
Configure Asset Text Indexing via local-appsettings.config
The asset text search feature requires system administrators to configure the asset content indexing settings in local-appsettings.config.
To enable asset text search for users, set the AssetContentIndexing_Enabled key to true in Drive:]\[path-to-cms-site-instance]\local-appsettings.config.
System administrators can also configure other asset text search settings to suit their organization's needs, including side-by-side indexing and search optimization cycles.
See Configuring local-appsettings.config for steps to configure local-appsettings.config, and see Key/Value Pairs in local-appsettings.config for details about available asset text search settings.
Asset Search Index Location
The CMS generates the asset search index in the following folder location:
Drive:]\[path-to-cms-site-instance]\site\App_Data\AssetSearchIndex
Asset Text Search Ranking Logic
When asset text search functionality is enabled and users run searches on assets, the search results list displays text-based entries at the top. This behavior occurs no matter if pages and components are also included in the query.
For search ranking scores, the CMS uses the standard search algorithm instead of boosts for the document name, title, and full text content. This algorithm does not apply to ID and hierarchy. Also, asset text searches do not override term frequency (TF) and term frequency under the context of the document content (IDF).
See how this logic plays into asset text searches via the following example.
Let's say a user searches for the text "garage" within assets. In the results list, an asset has a five star ranking while another asset has a one star ranking. Within the five star asset, the text "garage" displays only once, and the full text content is relatively small. Within the one star asset, the text "garage" displays twice, and the full text content is extensive.
Why does the asset with two instances of "garage" receive a lower ranking than the asset with only one? This occurs because the five star asset has a 50% IDF score compared to the size of the document.