Fix small error in the new doc section

Author: StroemPhi

docs/automatic-generation.md

@@ -91,7 +91,7 @@ We apply three main strategies to these underspecified classes:
 
 Some classes require more than just generic identifiers, labels or free-text descriptions to be useful. Rather than inventing new properties, we reuse slots already specified in the source ontologies these classes originate from:
 
-  * `Location`: Defined in the [Location Core Vocabulary (LOCN)](https://semiceu.github.io/Core-Location-Vocabulary/releases/2.1.1/). We add slots like `bbox`, `centroid`, and `geometry` (with sub-slots for `wkt`, `gml`, `latitude`, `longitude`). These are standard LOCN properties, ensuring compatibility with existing geospatial tools. 
+  * `Geometry`: Defined in the [Location Core Vocabulary (LOCN)](https://semiceu.github.io/Core-Location-Vocabulary/releases/2.1.1/). We add slots like `wkt`, `gml`, `latitude`, `longitude`). These are standard LOCN properties, ensuring compatibility with existing geospatial tools. 
   * `Kind`: Defined in the [vCard Ontology](https://www.w3.org/TR/vcard-rdf/) and used as the range for the `contact_point` slot. We add `formatted_name`, `has_email`, and `has_telephone`. These are direct mappings to vCard properties (`vcard:fn`, `vcard:hasEmail`, `vcard:hasTelephone`), ensuring contact information is machine-actionable according to the standard. 
   * `Agent`: In addition to the existing `name` slot, we add `identifier` and `other_identifier` slots. This allows for direct disambiguation of agents (e.g., using an ORCID for a person or a ROR ID for an organization), which is crucial for data integration across different repositories. We also add `rdf_type` form the DCAT-AP+ [ClassifierMixin](design-patterns.md#pattern-3-flexible-classification-classifiermixin) to allow explicit additional typing with `foaf:Person` and `foaf:Organization`.