Extraction

Named Entity Recognition extraction is the process of matching and retrieving entities from text, according to the rules in your grammars.

To run Named Entity Recognition, you send a text file, or the raw text, to the Named Entity Recognition engine (you can do this in several ways, see Named Entity Recognition Architecture). You use the Named Entity Recognition configuration to specify the grammars and entities that you want to match. Named Entity Recognitionidentifies each instance of the requested entity, and returns an XML list of matches.

By default, Named Entity Recognition returns the matched entity, with some additional information about the match, such as a confidence score for the accuracy of the match. It can also identify any configured components of the entity match, such as the parts of a social security number or phone number (see Components).