Identifying People, Places, and Organizations
Use a linguistic tagger to perform named entity recognition on a string.
Overview
Identifying named entities in natural language text can help make your app more intelligent. For example, a messaging app might look for names of people and places in text in order to display related information like contact information or directions.
The example below shows how to use NSLinguisticTagger to enumerate over natural language text and identify any named person, place, or organization.
let text = "The American Red Cross was established in Washington, D.C., by Clara Barton."
let tagger = NSLinguisticTagger(tagSchemes: [.nameType], options: 0)
tagger.string = text
let range = NSRange(location:0, length: text.utf16.count)
let options: NSLinguisticTagger.Options = [.omitPunctuation, .omitWhitespace, .joinNames]
let tags: [NSLinguisticTag] = [.personalName, .placeName, .organizationName]
tagger.enumerateTags(in: range, unit: .word, scheme: .nameType, options: options) { tag, tokenRange, stop in
if let tag = tag, tags.contains(tag) {
let name = (text as NSString).substring(with: tokenRange)
print("\(name): \(tag)")
}
}First, an instance of NSLinguisticTagger is created, specifying nameType as the tag scheme to be used. Next, the string property of the linguistic tagger is set to the natural language text. Finally, the linguistic tagger enumerates over the entire range of the string, specifying NSLinguisticTaggerUnit.word as the tag unit and nameType as the tag scheme, omitting any punctuation or whitespace, and joining words that are part of a single name into the same token. In the enumeration block, the name type is provided by tag, and each word is obtained by taking a substring of the original text at tokenRange.
When run, this code prints out each name and its type on a new line, as shown below:
Name | Type |
|---|---|
The American Red Cross | |
Washington, D.C. | |
Clara Barton |