How to extract hidden information and ‘aboutness’ from text using SKOS, ontologies, corpus analysis and linked data

Automatic annotation and classification of texts are often performed using machine learning techniques, statistical methods, controlled vocabularies and ontologies (e.g. SKOS), or comprehensive knowledge graphs such as DBpedia or Wikidata.

All these methods have advantages and disadvantages as discussed and summarized in this presentation. But what if all these methods are combined? What new possibilities for text analysis arise from this, and how does this influence the governance model underlying the knowledge modeling process?