Example Use Case
To illustrate the workflow in practice, this example follows the processing of a reception document related to a translated Dutch literary work, namely Marieke Lucas Rijneveld’s De avond is ongemak (The Discomfort of Evening).
When the document is uploaded to the DLBT, the system first applies OCR correction to ensure text quality and detect key entities such as names, publication dates, and referenced works. In the first AI stage, the model automatically generates an iConTxtInfo package containing an English-language translation, summary, central research question, and key thematic descriptors. This structured information becomes part of the iConTxt knowledge base, linking the document to related entries such as the author, translator, and publisher.
In the second stage, the RAG pipeline combines this internal knowledge base with external data sources to produce a new contextual text. This text provides a synthesized overview of Rijneveld’s position within the Dutch literary field and her reception abroad, referencing relevant translations, publishers, and prizes.