Share via


Text Analysis Authoring - Get Model Evaluation Summary

トレーニング済みモデルの評価の概要を取得します。 概要には、F 値、適合率、再現率など、モデルの高レベルのパフォーマンス測定が含まれます。

GET {Endpoint}/language/authoring/analyze-text/projects/{projectName}/models/{trainedModelLabel}/evaluation/summary-result?api-version=2023-04-01

URI パラメーター

名前 / 必須 説明
Endpoint
path True

string

サポートされている Cognitive Services エンドポイント (例: https://.api.cognitiveservices.azure.com).

projectName
path True

string

使用するプロジェクトの名前。

trainedModelLabel
path True

string

トレーニング済みのモデル ラベル。

api-version
query True

string

クライアント API バージョン。

応答

名前 説明
200 OK TextAnalysisAuthoringEvaluationSummary:

すべての評価結果の一覧。

Other Status Codes

ErrorResponse

エラー応答。

セキュリティ

AADToken

これらは Azure Active Directory OAuth2 フローです。 Azure ロールベースのアクセス制御と組み合わせて使用すると、Azure Maps REST API へのアクセスを制御できます。 Azure ロールベースのアクセス制御は、1 つ以上のAzure Mapsリソース アカウントまたはサブリソースへのアクセスを指定するために使用されます。 REST API をAzure Mapsするための 1 つ以上のアクセス許可で構成される組み込みロールまたはカスタム ロールを使用して、ユーザー、グループ、またはサービス プリンシパルにアクセス権を付与できます。

シナリオを実装するには、 認証の概念を表示することをお勧めします。 要約すると、このセキュリティ定義は、特定の API とスコープに対するアクセス制御が可能なオブジェクトを使用してアプリケーションをモデル化するためのソリューションを提供します。

メモ

  • このセキュリティ定義では、 ヘッダーを使用して、x-ms-client-idアプリケーションがアクセスを要求しているリソースAzure Maps示す必要があります。 これは、 Maps 管理 API から取得できます。

Authorization URL 、Azure パブリック クラウド インスタンスに固有です。 ソブリン クラウドには、一意の承認 URL と Azure Active Directory 構成があります。 * Azure ロールベースのアクセス制御は、Azure portal、PowerShell、CLI、Azure SDK、または REST API を介して Azure 管理プレーンから構成されます。 * Azure Maps Web SDK を使用すると、複数のユース ケースに対してアプリケーションを構成ベースで設定できます。

  • 現在、Azure Active Directory v1.0 または v2.0 では、職場、学校、ゲストはサポートされていますが、個人用アカウントはサポートされていません。

Type: oauth2
Flow: implicit
Authorization URL: https://login.microsoftonline.com/common/oauth2/authorize

Scopes

名前 説明
https://cognitiveservices.azure.com/.default https://cognitiveservices.azure.com/.default

Ocp-Apim-Subscription-Key

言語サービス リソースのサブスクリプション キー。

Type: apiKey
In: header

Successful Get Model Evaluation Summary

Sample Request

GET {Endpoint}/language/authoring/analyze-text/projects/LoanAgreements/models/model2/evaluation/summary-result?api-version=2023-04-01

Sample Response

{
  "projectKind": "CustomEntityRecognition",
  "customEntityRecognitionEvaluation": {
    "confusionMatrix": {
      "BorrowerAddress": {
        "BorrowerAddress": {
          "normalizedValue": 86.206894,
          "rawValue": 3.4482758
        },
        "$none": {
          "normalizedValue": 13.793103,
          "rawValue": 0.55172414
        }
      },
      "BorrowerCity": {
        "BorrowerCity": {
          "normalizedValue": 100,
          "rawValue": 4
        }
      },
      "BorrowerName": {
        "BorrowerName": {
          "normalizedValue": 100,
          "rawValue": 4
        }
      },
      "BorrowerState": {
        "BorrowerState": {
          "normalizedValue": 100,
          "rawValue": 4
        }
      },
      "Date": {
        "Date": {
          "normalizedValue": 100,
          "rawValue": 4
        }
      },
      "Interest": {
        "Interest": {
          "normalizedValue": 100,
          "rawValue": 4
        }
      },
      "LenderAddress": {
        "LenderAddress": {
          "normalizedValue": 100,
          "rawValue": 4
        }
      },
      "LenderCity": {
        "LenderCity": {
          "normalizedValue": 100,
          "rawValue": 4
        }
      },
      "LenderName": {
        "LenderName": {
          "normalizedValue": 100,
          "rawValue": 4
        }
      },
      "LenderState": {
        "LenderState": {
          "normalizedValue": 100,
          "rawValue": 4
        }
      },
      "LoanAmountNumbers": {
        "LoanAmountNumbers": {
          "normalizedValue": 100,
          "rawValue": 4
        }
      },
      "LoanAmountWords": {
        "LoanAmountWords": {
          "normalizedValue": 100,
          "rawValue": 4
        }
      },
      "$none": {
        "$none": {
          "normalizedValue": 99.81485,
          "rawValue": 51.90372
        },
        "BorrowerAddress": {
          "normalizedValue": 0.18315019,
          "rawValue": 0.0952381
        },
        "Interest": {
          "normalizedValue": 0.002005294,
          "rawValue": 0.0010427529
        }
      }
    },
    "entities": {
      "Date": {
        "f1": 1,
        "precision": 1,
        "recall": 1,
        "truePositiveCount": 4,
        "trueNegativeCount": 0,
        "falsePositiveCount": 0,
        "falseNegativeCount": 0
      },
      "BorrowerName": {
        "f1": 1,
        "precision": 1,
        "recall": 1,
        "truePositiveCount": 4,
        "trueNegativeCount": 0,
        "falsePositiveCount": 0,
        "falseNegativeCount": 0
      },
      "BorrowerAddress": {
        "f1": 0.6666666865348816,
        "precision": 0.6000000238418579,
        "recall": 0.75,
        "truePositiveCount": 3,
        "trueNegativeCount": 0,
        "falsePositiveCount": 2,
        "falseNegativeCount": 1
      },
      "BorrowerCity": {
        "f1": 1,
        "precision": 1,
        "recall": 1,
        "truePositiveCount": 4,
        "trueNegativeCount": 0,
        "falsePositiveCount": 0,
        "falseNegativeCount": 0
      },
      "BorrowerState": {
        "f1": 1,
        "precision": 1,
        "recall": 1,
        "truePositiveCount": 4,
        "trueNegativeCount": 0,
        "falsePositiveCount": 0,
        "falseNegativeCount": 0
      },
      "LenderName": {
        "f1": 1,
        "precision": 1,
        "recall": 1,
        "truePositiveCount": 4,
        "trueNegativeCount": 0,
        "falsePositiveCount": 0,
        "falseNegativeCount": 0
      },
      "LenderAddress": {
        "f1": 1,
        "precision": 1,
        "recall": 1,
        "truePositiveCount": 4,
        "trueNegativeCount": 0,
        "falsePositiveCount": 0,
        "falseNegativeCount": 0
      },
      "LenderCity": {
        "f1": 1,
        "precision": 1,
        "recall": 1,
        "truePositiveCount": 4,
        "trueNegativeCount": 0,
        "falsePositiveCount": 0,
        "falseNegativeCount": 0
      },
      "LenderState": {
        "f1": 1,
        "precision": 1,
        "recall": 1,
        "truePositiveCount": 4,
        "trueNegativeCount": 0,
        "falsePositiveCount": 0,
        "falseNegativeCount": 0
      },
      "LoanAmountWords": {
        "f1": 1,
        "precision": 1,
        "recall": 1,
        "truePositiveCount": 4,
        "trueNegativeCount": 0,
        "falsePositiveCount": 0,
        "falseNegativeCount": 0
      },
      "LoanAmountNumbers": {
        "f1": 1,
        "precision": 1,
        "recall": 1,
        "truePositiveCount": 4,
        "trueNegativeCount": 0,
        "falsePositiveCount": 0,
        "falseNegativeCount": 0
      },
      "Interest": {
        "f1": 0.75,
        "precision": 0.75,
        "recall": 0.75,
        "truePositiveCount": 3,
        "trueNegativeCount": 0,
        "falsePositiveCount": 1,
        "falseNegativeCount": 1
      }
    },
    "microF1": 0.94845366,
    "microPrecision": 0.93877554,
    "microRecall": 0.9583333,
    "macroF1": 0.9513889,
    "macroPrecision": 0.9458334,
    "macroRecall": 0.9583333
  },
  "evaluationOptions": {
    "kind": "percentage",
    "trainingSplitPercentage": 80,
    "testingSplitPercentage": 20
  }
}

定義

名前 説明
Error

error オブジェクト。

ErrorCode

人間が判読できるエラー コード。

ErrorResponse

エラー応答。

EvaluationKind

評価の種類を表します。 既定では、評価の種類は percentage に設定されています。

InnerErrorCode

人間が判読できるエラー コード。

InnerErrorModel

エラーに関するより具体的な情報を含むオブジェクト。 Microsoft One API のガイドラインに従って - https://github.com/Microsoft/api-guidelines/blob/vNext/Guidelines.md#7102-error-condition-responses

ProjectKind

プロジェクトの種類。

TextAnalysisAuthoringCustomEntityRecognitionEvaluationSummary

カスタム エンティティ認識プロジェクトの評価の概要を表します。

TextAnalysisAuthoringCustomMultiLabelClassificationEvaluationSummary

カスタムマルチラベル分類プロジェクトの評価の概要を表します。

TextAnalysisAuthoringCustomSingleLabelClassificationEvaluationSummary

カスタムの単一ラベル分類プロジェクトの評価の概要を表します。

TextAnalysisAuthoringEntityEvaluationSummary

エンティティの評価の概要を表します。

TextAnalysisAuthoringEntityRecognitionEvaluationSummary

カスタム エンティティ認識プロジェクトの評価の概要を表します。

TextAnalysisAuthoringEvaluationOptions

評価の実行に使用されるオプションを表します。

TextAnalysisAuthoringMultiLabelClassEvaluationSummary

複数ラベル分類プロジェクトのクラスの評価の概要を表します。

TextAnalysisAuthoringMultiLabelClassificationEvaluationSummary

複数ラベル分類プロジェクトの評価の概要を表します。

TextAnalysisAuthoringSingleLabelClassEvaluationSummary

単一ラベル分類プロジェクトのクラスの評価の概要を表します。

TextAnalysisAuthoringSingleLabelClassificationEvaluationSummary

カスタムの単一ラベル分類プロジェクトの評価の概要を表します。

Error

error オブジェクト。

名前 説明
code

ErrorCode

サーバー定義のエラー コードのセットの 1 つ。

details

Error[]

この報告されたエラーの原因となった特定のエラーに関する詳細の配列。

innererror

InnerErrorModel

エラーに関する現在のオブジェクトよりも具体的な情報を含む オブジェクト。

message

string

エラーの人間が判読できる表現。

target

string

エラーのターゲット。

ErrorCode

人間が判読できるエラー コード。

名前 説明
AzureCognitiveSearchIndexLimitReached

string

AzureCognitiveSearchIndexNotFound

string

AzureCognitiveSearchNotFound

string

AzureCognitiveSearchThrottling

string

Conflict

string

Forbidden

string

InternalServerError

string

InvalidArgument

string

InvalidRequest

string

NotFound

string

OperationNotFound

string

ProjectNotFound

string

QuotaExceeded

string

ServiceUnavailable

string

Timeout

string

TooManyRequests

string

Unauthorized

string

Warning

string

ErrorResponse

エラー応答。

名前 説明
error

Error

error オブジェクト。

EvaluationKind

評価の種類を表します。 既定では、評価の種類は percentage に設定されています。

名前 説明
manual

string

データ内のすべての例について、選択したデータセットに従ってデータを分割します。

percentage

string

ユーザー定義のパーセンテージに従って、データをトレーニング セットとテスト セットに分割します。

InnerErrorCode

人間が判読できるエラー コード。

名前 説明
AzureCognitiveSearchNotFound

string

AzureCognitiveSearchThrottling

string

EmptyRequest

string

ExtractionFailure

string

InvalidCountryHint

string

InvalidDocument

string

InvalidDocumentBatch

string

InvalidParameterValue

string

InvalidRequest

string

InvalidRequestBodyFormat

string

KnowledgeBaseNotFound

string

MissingInputDocuments

string

ModelVersionIncorrect

string

UnsupportedLanguageCode

string

InnerErrorModel

エラーに関するより具体的な情報を含むオブジェクト。 Microsoft One API のガイドラインに従って - https://github.com/Microsoft/api-guidelines/blob/vNext/Guidelines.md#7102-error-condition-responses

名前 説明
code

InnerErrorCode

サーバー定義のエラー コードのセットの 1 つ。

details

object

エラーの詳細。

innererror

InnerErrorModel

エラーに関する現在のオブジェクトよりも具体的な情報を含む オブジェクト。

message

string

エラー メッセージ。

target

string

エラー ターゲット。

ProjectKind

プロジェクトの種類。

名前 説明
CustomEntityRecognition

string

独自のデータを使用してドメイン カテゴリを識別する抽出モデルを構築する場合。

CustomMultiLabelClassification

string

独自のデータを使用してテキストを分類する分類モデルを構築する場合。 各ファイルには、1 つまたは複数のラベルを含めることができます。 たとえば、ファイル 1 は A、B、C に分類され、ファイル 2 は B と C に分類されます。

CustomSingleLabelClassification

string

独自のデータを使用してテキストを分類する分類モデルを構築する場合。 各ファイルのラベルは 1 つだけです。 たとえば、ファイル 1 は A として分類され、ファイル 2 は B として分類されます。

TextAnalysisAuthoringCustomEntityRecognitionEvaluationSummary

カスタム エンティティ認識プロジェクトの評価の概要を表します。

名前 説明
customEntityRecognitionEvaluation

TextAnalysisAuthoringEntityRecognitionEvaluationSummary

抽出評価に関連するデータが含まれます。

evaluationOptions

TextAnalysisAuthoringEvaluationOptions

評価の実行に使用されるオプションを表します。

projectKind string:

CustomEntityRecognition

評価が実行されたプロジェクトの種類を表します。

TextAnalysisAuthoringCustomMultiLabelClassificationEvaluationSummary

カスタムマルチラベル分類プロジェクトの評価の概要を表します。

名前 説明
customMultiLabelClassificationEvaluation

TextAnalysisAuthoringMultiLabelClassificationEvaluationSummary

複数ラベル分類の評価に関連するデータが含まれます。

evaluationOptions

TextAnalysisAuthoringEvaluationOptions

評価の実行に使用されるオプションを表します。

projectKind string:

CustomMultiLabelClassification

評価が実行されたプロジェクトの種類を表します。

TextAnalysisAuthoringCustomSingleLabelClassificationEvaluationSummary

カスタムの単一ラベル分類プロジェクトの評価の概要を表します。

名前 説明
customSingleLabelClassificationEvaluation

TextAnalysisAuthoringSingleLabelClassificationEvaluationSummary

単一ラベル分類評価に関連するデータが含まれます。

evaluationOptions

TextAnalysisAuthoringEvaluationOptions

評価の実行に使用されるオプションを表します。

projectKind string:

CustomSingleLabelClassification

評価が実行されたプロジェクトの種類を表します。

TextAnalysisAuthoringEntityEvaluationSummary

エンティティの評価の概要を表します。

名前 説明
f1

number

モデルの精度を表します

falseNegativeCount

integer

偽陰性の数を表します

falsePositiveCount

integer

誤検知の数を表します

precision

number

モデルの再現率を表します

recall

number

モデル F1 スコアを表します

trueNegativeCount

integer

真負の数を表します

truePositiveCount

integer

真陽性の数を表します

TextAnalysisAuthoringEntityRecognitionEvaluationSummary

カスタム エンティティ認識プロジェクトの評価の概要を表します。

名前 説明
confusionMatrix

object

2 つのエンティティ間の混同行列を表します (2 つのエンティティは同じにすることができます)。 マトリックスは、ラベル付けされたエンティティと予測されたエンティティの間にあります。

entities

<string,  TextAnalysisAuthoringEntityEvaluationSummary>

エンティティの評価を表します

macroF1

number

マクロ F1 を表します

macroPrecision

number

マクロの有効桁数を表します

macroRecall

number

マクロの再現率を表します

microF1

number

マイクロ F1 を表します

microPrecision

number

マイクロ精度を表します

microRecall

number

マイクロリコールを表します

TextAnalysisAuthoringEvaluationOptions

評価の実行に使用されるオプションを表します。

名前 説明
kind

EvaluationKind

評価の種類を表します。 既定では、評価の種類は percentage に設定されています。

testingSplitPercentage

integer

テスト データセットの分割率を表します。 評価の種類がパーセンテージの場合にのみ必要です。

trainingSplitPercentage

integer

トレーニング データセットの分割率を表します。 評価の種類がパーセンテージの場合にのみ必要です。

TextAnalysisAuthoringMultiLabelClassEvaluationSummary

複数ラベル分類プロジェクトのクラスの評価の概要を表します。

名前 説明
f1

number

モデルの精度を表します

falseNegativeCount

integer

偽陰性の数を表します

falsePositiveCount

integer

誤検知の数を表します

precision

number

モデルの再現率を表します

recall

number

モデル F1 スコアを表します

trueNegativeCount

integer

真負の数を表します

truePositiveCount

integer

真陽性の数を表します

TextAnalysisAuthoringMultiLabelClassificationEvaluationSummary

複数ラベル分類プロジェクトの評価の概要を表します。

名前 説明
classes

<string,  TextAnalysisAuthoringMultiLabelClassEvaluationSummary>

クラスの評価を表します

macroF1

number

マクロ F1 を表します

macroPrecision

number

マクロの有効桁数を表します

macroRecall

number

マクロの再現率を表します

microF1

number

マイクロ F1 を表します

microPrecision

number

マイクロ精度を表します

microRecall

number

マイクロリコールを表します

TextAnalysisAuthoringSingleLabelClassEvaluationSummary

単一ラベル分類プロジェクトのクラスの評価の概要を表します。

名前 説明
f1

number

モデルの精度を表します

falseNegativeCount

integer

偽陰性の数を表します

falsePositiveCount

integer

誤検知の数を表します

precision

number

モデルの再現率を表します

recall

number

モデル F1 スコアを表します

trueNegativeCount

integer

真負の数を表します

truePositiveCount

integer

真陽性の数を表します

TextAnalysisAuthoringSingleLabelClassificationEvaluationSummary

カスタムの単一ラベル分類プロジェクトの評価の概要を表します。

名前 説明
classes

<string,  TextAnalysisAuthoringSingleLabelClassEvaluationSummary>

クラスの評価を表します

confusionMatrix

object

2 つのクラス間の混同行列を表します (2 つのクラスは同じにすることができます)。 マトリックスは、ラベル付けされたクラスと予測されたクラスの間にあります。

macroF1

number

マクロ F1 を表します

macroPrecision

number

マクロの有効桁数を表します

macroRecall

number

マクロの再現率を表します

microF1

number

マイクロ F1 を表します

microPrecision

number

マイクロ精度を表します

microRecall

number

マイクロリコールを表します