Named Entity Recognition

Displays the named entity extracted from the given text. The covered entities are:

• Person
• Location
• Organization

Statistics

Example Use Cases

The NER classifier helps reduce the noise in the query results. For example, assuming one is looking for the word “Budget” in the request to the API. In that case, we likely have results like:

• Example 1: I need to manage my budget.
• Example 2: I have just called Budget to reserve a sedan for the next week.

The budget in example 1 is a general word, while it is the rental car company in the second example. Provided that we are looking for Budget as an organization, example 1 would be a noise that can negatively impact the extracted results of the query. In this use case, the user can easily use the NER classifier to reduce the noise by asking the API to return sentences where the “Budget” keyword is extracted as an organization.

Compatible Data Scoures

As a stream-integrated classifier, it is run on ingestion for specific sources.

Output

The NER classifier can add one or multiple metadata sections in results on the extracted Person/Location/Organization. Unlike other classifiers, the NER classifier's output is in the form of a list to cover more than one entity. For instance, the location entities in the following example is made of two labels, while the person and organization entities have one.

"enrichment": 
            { 
	"location_entities": [ 
    { "label": "Germany" },
    { "label": "France" } 
                                 ],
 "person_entities": [ 
   { "label": "Putin" } ] },
 
 "organization_entities": [ 
   { "label": "Microsoft" } 
 ] 
}