Chapter 7: Entity Roles and Groups for Complex Named Entity Recognition

In Chapter 2, Natural Language Understanding in Rasa, we introduced how to carry out Named Entity Recognition (NER) in Rasa. NER extracts the entity type and the entity value from a piece of text. Unfortunately, for complex NER, we require more information than simply the entity type and the entity value. In this chapter, we will introduce the entity roles and entity groups for dealing with complex NER problems. The entity role can be used to distinguish the different semantic roles of entities (that have the same entity type). In comparison, the entity group can be used to group entities into different groups, where each grouped entity belongs to different subtasks in the same request.

In this chapter, you will learn how entity roles and entity groups can be used to solve the complex NER problem. Additionally, you will learn how to define training data, configure pipelines, and write stories for entity roles...