hermannkroll / kgextractiontoolbox Goto Github PK

Currently, the computation of the document classification may classify one document multiple times with the same class. That results in overhead that can be avoided. An additional table is necessary to store the documents with the classifications. The following steps are required for this:

1. add tabledoc_classified_for with the columns document_id, collection, classification, and date_inserted (as default)
2. change the script which is used for the document classification (classification.py)
   • force-parameter to ignore stored classifications
   • function to query already classified documents (equivalent to get_untagged_doc_ids_by_tagger)
   • functionality that utilizes the cached data
3. add a function to insert document classifications (equivalent to add_doc_tagged_by
4. additional tests to validate the new functionality
5. change the documentation (README_02_CLASSIFICATIONS) accordingly
6. modify the DB-diagram