Document classification API

A2iA has introduced DocumentReader V3.0, a document classification and data extraction toolkit that automatically processes paper documents that contain both handwritten and machine-printed information and classifies those documents based on both their geometry and content. The A2iA DocumentReader API can be integrated into existing enterprise content management, document management and knowledge management applications.

The company says the latest release of this API features a new user interface that allows documents to be classified not only at an individual page-level but also is able to classify a group of documents together (such as would be found in an envelope or paper file-folder). A2iA adds DocumentReader employs advanced artificial intelligence techniques, along with new geometric and content analyses, to analyze sets of pages found together. Because A2iA DocumentReader analyzes a group of documents simultaneously, it corroborates the data from these related sources, improving classification rates and enhancing recognition accuracy, the company claims.

A2iA DocumentReader is capable of reading free form, cursive handwriting from unstructured documents and can locate and extract key data elements such as social security numbers, phone numbers, dates and product names. Users can define their own set of keywords and phrases to be located, eliminating the time and cost associated with manually reviewing and sorting documents.