Discrepancy handler for document ingestion into a corpus for a cognitive computing system

A computer program product for ingesting a new document into a corpus of a cognitive computing system comprising:

• a tangible, computer-readable computer memory device which is not a propagating signal per se;
• and
• program instructions encoded by the computer memory device for causing a processor to, when executed, perform steps comprising:
• receiving at least a portion of the new document prior to ingestion of at least the portion of the new document into the corpus of the cognitive computing system;
• prior to ingestion of at least the portion of the new document into the corpus:
• automatically querying using the cognitive computing system, the corpus of the cognitive computing system plus at least the portion of the new document outside the corpus according to one or more questions previously associated with a category of documents in the corpus;
• responsive to the querying, automatically detecting one or more discrepancies between the portion and the corpus;
• assigning a confidence factor to each detected discrepancy;
• reporting the one or more discrepancies and the assigned confidence factors to a user;
• and
• responsive to a user command, ingesting the portion into the corpus.

A system for ingesting a new document into a corpus of a cognitive computing system comprising:

• a computing system having a processor;
• a tangible, computer-readable computer memory device which is not a propagating signal per se;
• and
• program instructions encoded by the computer memory device for causing a processor to, when executed, perform steps comprising:
• receiving at least a portion of the new document prior to ingestion of at least the portion of the new document into the corpus of the cognitive computing system;
• prior to ingestion of at least the portion of the new document into the corpus:
• automatically querying using the cognitive computing system, the corpus of the cognitive computing system plus at least the portion of the new document outside the corpus according to one or more questions previously associated with a category of documents in the corpus;
• responsive to the querying, automatically detecting one or more discrepancies between the portion and the corpus;
• assigning a confidence factor to each detected discrepancy;
• reporting the one or more discrepancies and the assigned confidence factors to a user;
• and
• responsive to a user command, ingesting the portion into the corpus.