natural-language（自然语言）

NaturalLanguage + Translation

Analyze natural language text for tokenization, part-of-speech tagging, named entity recognition, sentiment analysis, language identification, and word/sentence embeddings. Translate text between languages with the Translation framework. Targets Swift 6.3 / iOS 26+.

This skill covers two related frameworks: NaturalLanguage (NLTokenizer, NLTagger, NLEmbedding) for on-device text analysis, and Translation (TranslationSession, LanguageAvailability) for language translation.

Contents

• Setup
• Tokenization
• Language Identification
• Part-of-Speech Tagging
• Named Entity Recognition
• Sentiment Analysis
• Text Embeddings
• Translation
• Common Mistakes
• Review Checklist
• References

Setup

Import NaturalLanguage for text analysis and Translation for language translation. No special entitlements or capabilities are required for NaturalLanguage. Translation requires iOS 17.4+ / macOS 14.4+.

import NaturalLanguage
import Translation

NaturalLanguage classes (NLTokenizer, NLTagger) are not thread-safe. Use each instance from one thread or dispatch queue at a time.

Tokenization

Segment text into words, sentences, or paragraphs with NLTokenizer.

import NaturalLanguage

func tokenizeWords(in text: String) -> [String] {
    let tokenizer = NLTokenizer(unit: .word)
    tokenizer.string = text

    let range = text.startIndex..<text.endIndex
    return tokenizer.tokens(for: range).map { String(text[$0]) }
}

Token Units

Enumerating with Attributes

Use enumerateTokens(in:using:) to detect numeric or emoji tokens.

let tokenizer = NLTokenizer(unit: .word)
tokenizer.string = text

tokenizer.enumerateTokens(in: text.startIndex..<text.endIndex) { range, attributes in
    if attributes.contains(.numeric) {
        print("Number: \(text[range])")
    }
    return true // continue enumeration
}

Language Identification

Detect the dominant language of a string with NLLanguageRecognizer.

func detectLanguage(for text: String) -> NLLanguage? {
    NLLanguageRecognizer.dominantLanguage(for: text)
}

// Multiple hypotheses with confidence scores
func languageHypotheses(for text: String, max: Int = 5) -> [NLLanguage: Double] {
    let recognizer = NLLanguageRecognizer()
    recognizer.processString(text)
    return recognizer.languageHypotheses(withMaximum: max)
}

Constrain the recognizer to expected languages for better accuracy on short text.

let recognizer = NLLanguageRecognizer()
recognizer.languageConstraints = [.english, .french, .spanish]
recognizer.processString(text)
let detected = recognizer.dominantLanguage

Part-of-Speech Tagging

Identify nouns, verbs, adjectives, and other lexical classes with NLTagger.

func tagPartsOfSpeech(in text: String) -> [(String, NLTag)] {
    let tagger = NLTagger(tagSchemes: [.lexicalClass])
    tagger.string = text

    var results: [(String, NLTag)] = []
    let range = text.startIndex..<text.endIndex
    let options: NLTagger.Options = [.omitPunctuation, .omitWhitespace]

    tagger.enumerateTags(in: range, unit: .word, scheme: .lexicalClass, options: options) { tag, tokenRange in
        if let tag {
            results.append((String(text[tokenRange]), tag))
        }
        return true
    }
    return results
}

Common Tag Schemes

Named Entity Recognition

Extract people, places, and organizations.

func extractEntities(from text: String) -> [(String, NLTag)] {
    let tagger = NLTagger(tagSchemes: [.nameType])
    tagger.string = text

    var entities: [(String, NLTag)] = []
    let options: NLTagger.Options = [.omitPunctuation, .omitWhitespace, .joinNames]

    tagger.enumerateTags(
        in: text.startIndex..<text.endIndex,
        unit: .word,
        scheme: .nameType,
        options: options
    ) { tag, tokenRange in
        if let tag, tag != .other {
            entities.append((String(text[tokenRange]), tag))
        }
        return true
    }
    return entities
}
// NLTag values: .personalName, .placeName, .organizationName

Sentiment Analysis

Score text sentiment from -1.0 (negative) to +1.0 (positive).

func sentimentScore(for text: String) -> Double? {
    let tagger = NLTagger(tagSchemes: [.sentimentScore])
    tagger.string = text

    let (tag, _) = tagger.tag(
        at: text.startIndex,
        unit: .paragraph,
        scheme: .sentimentScore
    )
    return tag.flatMap { Double($0.rawValue) }
}

Text Embeddings

Measure semantic similarity between words or sentences with NLEmbedding.

func wordSimilarity(_ word1: String, _ word2: String) -> Double? {
    guard let embedding = NLEmbedding.wordEmbedding(for: .english) else { return nil }
    return embedding.distance(between: word1, and: word2, distanceType: .cosine)
}

func findSimilarWords(to word: String, count: Int = 5) -> [(String, Double)] {
    guard let embedding = NLEmbedding.wordEmbedding(for: .english) else { return [] }
    return embedding.neighbors(for: word, maximumCount: count, distanceType: .cosine)
}

Sentence embeddings compare entire sentences.

func sentenceSimilarity(_ s1: String, _ s2: String) -> Double? {
    guard let embedding = NLEmbedding.sentenceEmbedding(for: .english) else { return nil }
    return embedding.distance(between: s1, and: s2, distanceType: .cosine)
}

Translation

System Translation Overlay

Show the built-in translation UI with .translationPresentation().

import SwiftUI
import Translation

struct TranslatableView: View {
    @State private var showTranslation = false
    let text = "Hello, how are you?"

    var body: some View {
        Button { showTranslation = true } label: {
            Text(text)
        }
        .buttonStyle(.plain)
        .translationPresentation(
            isPresented: $showTranslation,
            text: text
        )
    }
}

Programmatic Translation

Use .translationTask() for programmatic translations within a view context.

struct TranslatingView: View {
    @State private var translatedText = ""
    @State private var configuration: TranslationSession.Configuration?

    var body: some View {
        VStack {
            Text(translatedText)
            Button("Translate") {
                configuration = .init(source: Locale.Language(identifier: "en"),
                                      target: Locale.Language(identifier: "es"))
            }
        }
        .translationTask(configuration) { session in
            let response = try await session.translate("Hello, world!")
            translatedText = response.targetText
        }
    }
}

Batch Translation

Translate multiple strings in a single session.

.translationTask(configuration) { session in
    let requests = texts.enumerated().map { index, text in
        TranslationSession.Request(sourceText: text,
                                    clientIdentifier: "\(index)")
    }
    let responses = try await session.translations(from: requests)
    for response in responses {
        print("\(response.sourceText) -> \(response.targetText)")
    }
}

Checking Language Availability

let availability = LanguageAvailability()
let status = await availability.status(
    from: Locale.Language(identifier: "en"),
    to: Locale.Language(identifier: "ja")
)
switch status {
case .installed: break    // Ready to translate offline
case .supported: break    // Needs download
case .unsupported: break  // Language pair not available
}

Common Mistakes

DON'T: Share NLTagger/NLTokenizer across threads

These classes are not thread-safe and will produce incorrect results or crash.

// WRONG
let sharedTagger = NLTagger(tagSchemes: [.lexicalClass])
DispatchQueue.concurrentPerform(iterations: 10) { _ in
    sharedTagger.string = someText  // Data race
}

// CORRECT
await withTaskGroup(of: Void.self) { group in
    for _ in 0..<10 {
        group.addTask {
            let tagger = NLTagger(tagSchemes: [.lexicalClass])
            tagger.string = someText
            // process...
        }
    }
}

DON'T: Confuse NaturalLanguage with Core ML

NaturalLanguage provides built-in linguistic analysis. Use Core ML for custom trained models. They complement each other via NLModel.

// WRONG: Trying to do NER with raw Core ML
let coreMLModel = try MLModel(contentsOf: modelURL)

// CORRECT: Use NLTagger for built-in NER
let tagger = NLTagger(tagSchemes: [.nameType])

// Or load a custom Core ML model via NLModel
let nlModel = try NLModel(mlModel: coreMLModel)
tagger.setModels([nlModel], forTagScheme: .nameType)

DON'T: Assume embeddings exist for all languages

Not all languages have word or sentence embeddings available on device.

// WRONG: Force unwrap
let embedding = NLEmbedding.wordEmbedding(for: .japanese)!

// CORRECT: Handle nil
guard let embedding = NLEmbedding.wordEmbedding(for: .japanese) else {
    // Embedding not available for this language
    return
}

DON'T: Create a new tagger per token

Creating and configuring a tagger is expensive. Reuse it for the same text.

// WRONG: New tagger per word
for word in words {
    let tagger = NLTagger(tagSchemes: [.lexicalClass])
    tagger.string = word
}

// CORRECT: Set string once, enumerate
let tagger = NLTagger(tagSchemes: [.lexicalClass])
tagger.string = fullText
tagger.enumerateTags(in: fullText.startIndex..<fullText.endIndex,
                     unit: .word, scheme: .lexicalClass, options: []) { tag, range in
    return true
}

DON'T: Ignore language hints for short text

Language detection on short strings (under ~20 characters) is unreliable. Set constraints or hints to improve accuracy.

// WRONG: Detect language of a single word
let lang = NLLanguageRecognizer.dominantLanguage(for: "chat")  // French or English?

// CORRECT: Provide context
let recognizer = NLLanguageRecognizer()
recognizer.languageHints = [.english: 0.8, .french: 0.2]
recognizer.processString("chat")

Review Checklist

• NLTokenizer and NLTagger instances used from a single thread
• Tagger created once per text, not per token
• Language detection uses constraints/hints for short text
• NLEmbedding availability checked before use (returns nil if unavailable)
• Translation LanguageAvailability checked before attempting translation
• .translationTask() used within a SwiftUI view hierarchy
• Batch translation uses clientIdentifier to match responses to requests
• Sentiment scores handled as optional (may return nil for unsupported languages)
• .joinNames option used with NER to keep multi-word names together
• Custom ML models loaded via NLModel, not raw Core ML

References

• Extended patterns (custom models, contextual embeddings, gazetteers): references/translation-patterns.md
• Natural Language framework
• NLTokenizer
• NLTagger
• NLEmbedding
• NLLanguageRecognizer
• Translation framework
• TranslationSession
• LanguageAvailability