Classification and extraction of metadata with Mindbreeze InSpire

When you are using the Fabasoft Cloud Enterprise or Superior Edition in conjunction with Mindbreeze InSpire, you can automatically classify documents and extract metadata. The powerful concept for the incoming classification of documents is particularly beneficial if it is applied on custom forms.

How does it work?

The following steps illustrate the overall functionality:

  1. During a learning phase, Mindbreeze InSpire is trained to apply the classification of documents. For the extraction of metadata from documents, regular expressions can be defined.
  2. When documents are classified in production mode (triggered either manually or automatically), metadata and classifications are requested from the Mindbreeze InSpire service. Feedback to the system continually improves the quality of classification.
  3. For the classification value, a category with the respective import ID is searched and assigned to the document.
  4. This category can be used to start a document-specific registration or to initialise a process. If metadata has been retrieved from the document, it is prefilled at registration and highlighted in colour.


Documents are classified automatically via the „inbox“. This can be achieved by defining a rule on the inbox which will execute the action “Classify With Mindbreeze InSpire” on all incoming elements.


In workflow, documents are classified and registered by starting an ad-hoc process with the activity “Register”. You can equally use this activity in a predefined BPMN process. In this case, however, you have to restrict the applicability of the BPMN process to documents to ensure that users can only start this process on documents.