Classification of documents

Document classification involves assigning one or more classes or categories to a document.

‍

Structured document

This is a document that follows a pattern. In this case, the layout and design (colors, fonts and images) are similar among the various document copies. An example of a structured document is the ID document where each individual copy has the same format.

‍

Semi-structured document

This is a document that has the characteristic of containing a certain type of information known in advance, which can change position and format within the document itself. The most classic example is an invoice.

‍

Unstructured document

This is a document that does not follow any constraints in terms of format or content. A concrete example of an unstructured document is a contract.

‍

Human error

These are unplanned errors caused by humans. These errors reduce the quality, safety, and accuracy of a given activity.

Extraction of information

Information extraction involves locating and retrieving data from documents and other data sources.

‍

Human-in-the-loop (HITL)

The process by which human experts verify results obtained from machine learning models. Combining Machine Learning with human review provides the best of both worlds.

‍

Intelligent document processing (IDP)

A type of document automation that involves the acquisition, extraction and processing of data from a variety of document formats. For the purpose of classifying, categorizing, and extracting relevant information, using artificial intelligence and deep learning technologies such as natural language processing (NLP) and computer vision techniques.

‍

Optical character recognition (OCR)

Technology that can detect the characters in a document and transfer them into machine-readable digital text.

‍

‍

‍

‍