Skip to content

Entities

What are Entities?

The myBiros Platform provides a series of use cases for extracting information from structured, semi-structured, and unstructured documents. For each use case offered in the Platform's catalog, a series of fields of interest have been defined that will be extracted by the artificial intelligence model.

An Entity is defined as the representation of the semantics of a field to which one or more occurrences of the information extracted from the document are associated.
For example, having defined the field "Name" a corresponding entity will be defined, containing all occurrences of the name within the document page.

Types of Entities

Depending on the method and level of abstraction of the extracted information, two types of entities are defined: simple entities and aggregated entities.

Simple entities

Simple entities refer to individual and autonomous units of data, to which a well-defined semantics has been associated, that do not depend on other entities for their definition or existence.

Aggregated entities

Aggregated entities are entities that consist of multiple simple entities. They represent a more complex structure and arise from the need to relate multiple simple entities to each other.

Simple and Aggregated Entities: A practical example

Let's consider a purchase order as an example. We described simple entities as individual pieces of information, each representing a distinct unit of data with a specific meaning. In the context of a purchase order, examples of simple entities include:

  • The customer name, which identifies the person making the purchase.
  • The order reference code, which uniquely identifies the transaction.
  • The product ID, which specifies the item being purchased.
  • The product price, which indicates the cost of a single unit.
  • The product quantity, which shows how many units of the product were ordered.

Each of these simple entities is independent and does not depend on other data for its definition.

In contrast, an aggregated entity might be a line item on the order. A line item is a more complex structure that combines several simple entities. For example, a line item will include the product ID, price, and quantity, all of which are necessary to fully describe an ordered item. The line item represents the relationship between these simple entities, aggregating them into one entry that reflects the product’s role in the transaction. Thus, the line item is an aggregated entity because it consolidates multiple simple entities (product ID, price, quantity) into a single, cohesive representation of a product in the order.