Named entity recognition (NER) is a process of finding instances of specified entities in text. For example, a simple news named-entity recognizer for the English language might find the person "John J. Smith" and the location "Seattle" in the text "John J. Smith lives in Seattle".
NER functions let you specify how to extract entities when training the data models. The Aster Analytics Foundation provides two sets of NER functions.
The NER functions that use the Max Entropy Model model are:
- TrainNamedEntityFinder, which takes training data and outputs a Max Entropy data model
-
FindNamedEntity, which takes input documents and extracts specified entities, using a Max Entropy model and, if appropriate, rules (regular expressions) or a dictionary
The function uses a model to extract the entity types "person", "location", and "organization" and rules to extract the entity types "date", "time", "email" and "money". If you specify these entity names, the function invokes the default model types and model file names. To extract all entities in one FindNamedEntity call, specify "all".
- Evaluate Named Entity Finder, which evaluates a Max Entropy model
The maximum entropy model implementation supports only English text. The CRF model implementation supports English, simplified Chinese, and traditional Chinese text.
The NER functions that use the Conditional Random Fields (CRF) Model are documented in NER Functions (CRF Model Implementation).