Agency notice processing system

An artificial intelligence training system comprising:

• a computer system;
• a data loader in the computer system, wherein the data loader retrieves historical data that includes historical text generated from optical character recognition performed on historical images of agency notices and historical metadata for items in the historical images of the agency notices and retrieves validation data, wherein the validation data includes validation text generated from optical character recognition performed on validation images of additional agency notices and validation metadata for the items in the validation images of the additional agency notices;
• an annotator in the computer system, wherein the annotator generates annotations for the historical text in which the annotations identify items and locations of the items in the historical text;
• a trainer in the computer system, wherein trainer trains the artificial intelligence system using the historical data and the annotations;
• and
• a validator in the computer system, wherein the validator sends validation text to the artificial intelligence system, receives results from the artificial intelligence system processing the validation text, and determines whether the artificial intelligence system identifies the items in the validation images within a desired level of accuracy.

A computer program product for training an artificial intelligence system to process agency notices, the computer program product comprising:

• a computer-readable-storage media with first program code, second program code, and third computer-readable storage media;
• first program code, stored on the computer-readable-storage media, for identifying historical data that includes historical text generated from optical character recognition performed on historical images of the agency notices and historical metadata for items in the historical images of the agency notices;
• second program code, stored on the computer-readable-storage media, for generating annotations for the historical text, wherein the annotations identify items in the historical text and locations of the items in the historical text;
• and
• third program code, stored on the computer-readable-storage media, for training the artificial intelligence system using the historical data and the annotations.