Does anyone know the steps involved to make AEM Assets defined taxonomy (hierarchical tags) available in Content Hub? (as filters, upload metadata, and search)
Although some tags related choices show in the selector, these are not our custom defined ones. The configuration panel is very well made and easy to understand. But it is somewhat surprising that this essential part is not directly available.
Indeed this is not available as of now. The system automatically retrieves tags from the published assets and show in the Tags section. It is not possible to control of what it shows and what not. Since Content Hub is still evolving, I hope they let us configure Tags in future.
AEM Assets taxonomy (tags) can be made available in Content Hub when assets are synchronized via the Content Hub connector. Tags are transferred as asset metadata (commonly mapped to a “tags/keywords” field), searchable in Content Hub, and can be surfaced on downstream systems if your mapping and sync rules are configured correctly.
Common limitations and considerations
Mapping required: Tags are not magically exposed without mapping; connector configs must explicitly map AEM tag properties.
Hierarchy and namespaces: Some tag hierarchies or namespaces may be flattened or need custom transforms to preserve structure.
Localization: Tag localization may not automatically carry across; you may need custom handling.
Real-time updates: Depending on your sync configuration, tag updates in AEM might not appear instantly in Content Hub—expect sync windows or manual re-sync.
Search/indexing: Content Hub search depends on how metadata is indexed—ensure the mapped field is indexed for full-text or faceted search if needed.
Versioning / provenance: Content Hub stores metadata at ingest; if tags change frequently, consider how you want updates to propagate and whether to keep provenance.
Compatibility: Confirm your Content Hub / AEM versions support the desired mapping features—older connector versions may have limitations.
The AEM Content Hub cloud service connection in AEM (Tools → Cloud Services → Content Hub or the Cloud Services configuration you use). This holds the tenant ID, client credentials and endpoint being used.
The authentication/integration settings (Adobe I/O or OAuth client credentials) used to authenticate AEM to your Content Hub tenant.
Connector-specific mapping and sync settings (the metadata mapping rules, scheduled sync options, batch size, and filters) that live either in the AEM connector configuration or in the Content Hub mapping UI/ingest settings.
Runtime OSGi settings in AEM (visible in /system/console/configMgr) that can affect connector behavior (timeouts, retry, logging) if you need to troubleshoot.
Tenant-side Content Hub admin settings (metadata schemas, ingestion/transformation rules, indexing options) that determine how incoming metadata is stored and indexed.
“Custom transforms to preserve structure” — what I meant and how to do it
Clarification: I meant custom metadata transformation logic, not “custom login.” If you don’t want tag hierarchy flattened, you must transform the AEM tag structure into a form Content Hub can natively represent.
Typical challenge: AEM tags use paths/namespaces (e.g., /etc/tags/marketing/campaigns/summer) or namespaced keys (namespace:tag). Many ingestion pipelines simply map tags to a flat “keywords” array, losing hierarchy.
Options to preserve structure:
Send full paths or namespace-qualified values as a single string (e.g., "marketing/campaigns/summer") so the path is preserved in one field.
Create structured metadata (level1, level2, level3 fields) by parsing the tag path and populating multiple fields during ingestion.
Populate a custom JSON field that stores the hierarchical tree; downstream apps can read the structure.
Map AEM taxonomy to a matching hierarchical taxonomy in Content Hub (if Content Hub supports hierarchical taxonomies) rather than mapping to a flat keyword field.
Where to implement transforms:
On the AEM side before sync: use connector mapping transforms, a preprocessing workflow step, or a metadata processor to reshape tag values.
On the Content Hub side during ingestion: use the tenant’s ingestion/mapping rules or scripting (if supported) to parse incoming tag values into the desired structure.
Can we control indexing in Content Hub?
Short answer: Yes — to an extent.
What you can typically control:
Which metadata fields are indexed or marked searchable.
Which fields are facetable for filters and which are stored/displayed.
Field types (string, date, enum) and whether a field participates in full-text vs. exact-match searches.
Mapping of incoming metadata to those indexed fields.
Limits to be aware of:
In hosted SaaS Content Hub you generally cannot change low-level search engine internals (custom analyzers, tokenizers, or the physical index engine configuration) — those are managed by the service.
The exact UI/controls and granularity depend on your Content Hub version/tenant—some provide richer schema/index settings, others are more opinionated.
Practical path:
Define the metadata schema in Content Hub and mark required fields as indexed/searchable/facetable.
Ensure your AEM→Content Hub mapping targets those fields.
If you need advanced search behavior (stemming, synonyms, special analyzers) confirm with Content Hub support or your tenant admin whether those are configurable or require a feature request.
Why? Content Hub exposes schema/index flags in the admin area and the connector mapping aligns incoming data to those schema fields. But low-level search engine customizations are generally restricted on SaaS platforms.
