在 iOS 上使用 ML Kit 辨識文字語言

您可以使用ML Kit來識別文字字串的語言。您可以獲得字串最可能的語言,或取得字串所有可能語言的置信度分數。

ML Kit 可辨識 103 種不同語言的本機腳本文字。此外,還可以辨識阿拉伯語、保加利亞語、中文、希臘語、印地語、日語和俄語的羅馬化文本。

在你開始之前

  1. 如果您尚未將 Firebase 新增至您的應用程式中,請按照入門指南中的步驟進行操作。
  2. 在 Podfile 中包含 ML Kit 函式庫:
    pod 'Firebase/MLNaturalLanguage', '6.25.0'
    pod 'Firebase/MLNLLanguageID', '6.25.0'
    
    安裝或更新專案的 Pod 後,請務必使用其.xcworkspace開啟 Xcode 專案。
  3. 在您的應用程式中,導入 Firebase:

    迅速

    import Firebase

    Objective-C

    @import Firebase;

識別字串的語言

若要辨識字串的語言,請取得LanguageIdentification的實例,然後將字串傳遞給identifyLanguage(for:)方法。

例如:

迅速

let languageId = NaturalLanguage.naturalLanguage().languageIdentification()

languageId.identifyLanguage(for: text) { (languageCode, error) in
  if let error = error {
    print("Failed with error: \(error)")
    return
  }
  if let languageCode = languageCode, languageCode != "und" {
    print("Identified Language: \(languageCode)")
  } else {
    print("No language was identified")
  }
}

Objective-C

FIRNaturalLanguage *naturalLanguage = [FIRNaturalLanguage naturalLanguage];
FIRLanguageIdentification *languageId = [naturalLanguage languageIdentification];

[languageId identifyLanguageForText:text
                         completion:^(NSString * _Nullable languageCode,
                                      NSError * _Nullable error) {
                           if (error != nil) {
                             NSLog(@"Failed with error: %@", error.localizedDescription);
                             return;
                           }
                           if (languageCode != nil
                               && ![languageCode isEqualToString:@"und"] ) {
                             NSLog(@"Identified Language: %@", languageCode);
                           } else {
                             NSLog(@"No language was identified");
                           }
                         }];

如果呼叫成功,則會將BCP-47 語言代碼傳遞給完成處理程序,指示文字的語言。請參閱支援的語言的完整清單。如果無法可靠地偵測到任何語言,則通過代碼und (未確定)。

預設情況下,ML Kit 僅在識別置信度值至少為 0.5 的語言時才傳回非und值。您可以透過將LanguageIdentificationOptions物件傳遞給languageIdentification(options:)來變更此閾值:

迅速

let options = LanguageIdentificationOptions(confidenceThreshold: 0.4)
let languageId = NaturalLanguage.naturalLanguage().languageIdentification(options: options)

Objective-C

FIRNaturalLanguage *naturalLanguage = [FIRNaturalLanguage naturalLanguage];
FIRLanguageIdentificationOptions *options =
    [[FIRLanguageIdentificationOptions alloc] initWithConfidenceThreshold:0.4];
FIRLanguageIdentification *languageId =
    [naturalLanguage languageIdentificationWithOptions:options];

取得字串可能的語言

若要取得字串最可能的語言的置信度值,請取得LanguageIdentification的實例,然後將字串傳遞給identifyPossibleLanguages(for:)方法。

例如:

迅速

let languageId = NaturalLanguage.naturalLanguage().languageIdentification()

languageId.identifyPossibleLanguages(for: text) { (identifiedLanguages, error) in
  if let error = error {
    print("Failed with error: \(error)")
    return
  }
  guard let identifiedLanguages = identifiedLanguages,
    !identifiedLanguages.isEmpty,
    identifiedLanguages[0].languageCode != "und"
  else {
    print("No language was identified")
    return
  }

  print("Identified Languages:\n" +
    identifiedLanguages.map {
      String(format: "(%@, %.2f)", $0.languageCode, $0.confidence)
      }.joined(separator: "\n"))
}

Objective-C

FIRNaturalLanguage *naturalLanguage = [FIRNaturalLanguage naturalLanguage];
FIRLanguageIdentification *languageId = [naturalLanguage languageIdentification];

[languageId identifyPossibleLanguagesForText:text
                                  completion:^(NSArray<FIRIdentifiedLanguage *> * _Nonnull identifiedLanguages,
                                               NSError * _Nullable error) {
  if (error != nil) {
    NSLog(@"Failed with error: %@", error.localizedDescription);
    return;
  }
  if (identifiedLanguages.count == 1
      && [identifiedLanguages[0].languageCode isEqualToString:@"und"] ) {
    NSLog(@"No language was identified");
    return;
  }
  NSMutableString *outputText = [NSMutableString stringWithFormat:@"Identified Languages:"];
  for (FIRIdentifiedLanguage *language in identifiedLanguages) {
    [outputText appendFormat:@"\n(%@, %.2f)", language.languageCode, language.confidence];
  }
  NSLog(outputText);
}];

如果呼叫成功,則將IdentifiedLanguage物件清單傳遞給繼續處理程序。從每個物件中,您可以獲得該語言的 BCP-47 程式碼以及該字串採用該語言的置信度。請參閱支援的語言的完整清單。請注意,這些值表示整個字串採用給定語言的置信度; ML Kit 不會在單一字串中辨識多種語言。

預設情況下,ML Kit 僅傳回置信度值至少為 0.01 的語言。您可以透過將LanguageIdentificationOptions物件傳遞給languageIdentification(options:)來變更此閾值:

迅速

let options = LanguageIdentificationOptions(confidenceThreshold: 0.4)
let languageId = NaturalLanguage.naturalLanguage().languageIdentification(options: options)

Objective-C

FIRNaturalLanguage *naturalLanguage = [FIRNaturalLanguage naturalLanguage];
FIRLanguageIdentificationOptions *options =
    [[FIRLanguageIdentificationOptions alloc] initWithConfidenceThreshold:0.4];
FIRLanguageIdentification *languageId =
    [naturalLanguage languageIdentificationWithOptions:options];

如果沒有語言滿足此閾值,則清單將包含一項,其值為und