評価指標: 応答の類似性
上級
これはEngineering, AI分野の自動化ワークフローで、21個のノードを含みます。主にSet, Code, Merge, SplitOut, Aggregateなどのノードを使用、AI技術を活用したスマート自動化を実現。 評価指標: 応答の類似性
前提条件
- •ターゲットAPIの認証情報が必要な場合あり
- •OpenAI API Key
使用ノード (21)
ワークフロープレビュー
ノード接続関係を可視化、ズームとパンをサポート
ワークフローをエクスポート
以下のJSON設定をn8nにインポートして、このワークフローを使用できます
{
"meta": {
"instanceId": "408f9fb9940c3cb18ffdef0e0150fe342d6e655c3a9fac21f0f644e8bedabcd9"
},
"nodes": [
{
"id": "c5365531-2d66-4e67-9db2-37059e9e97a3",
"name": "OpenAI Chat Model1",
"type": "@n8n/n8n-nodes-langchain.lmChatOpenAi",
"position": [
-1512,
420
],
"parameters": {
"model": {
"__rl": true,
"mode": "list",
"value": "gpt-4.1-mini",
"cachedResultName": "gpt-4.1-mini"
},
"options": {}
},
"credentials": {
"openAiApi": {
"id": "8gccIjcuf3gvaoEr",
"name": "OpenAi account"
}
},
"typeVersion": 1.2
},
{
"id": "3881b831-b575-4b2f-8ebb-ac956a714f10",
"name": "データセット行取得時",
"type": "n8n-nodes-base.evaluationTrigger",
"position": [
-2040,
100
],
"parameters": {
"sheetName": {
"__rl": true,
"mode": "list",
"value": 1176192270,
"cachedResultUrl": "https://docs.google.com/spreadsheets/d/1YOnu2JJjlxd787AuYcg-wKbkjyjyZFgASYVV0jsij5Y/edit#gid=1176192270",
"cachedResultName": "Similarity"
},
"documentId": {
"__rl": true,
"mode": "list",
"value": "1YOnu2JJjlxd787AuYcg-wKbkjyjyZFgASYVV0jsij5Y",
"cachedResultUrl": "https://docs.google.com/spreadsheets/d/1YOnu2JJjlxd787AuYcg-wKbkjyjyZFgASYVV0jsij5Y/edit?usp=drivesdk",
"cachedResultName": "96. Evaluations Test"
}
},
"credentials": {
"googleSheetsOAuth2Api": {
"id": "XHvC7jIRR8A2TlUl",
"name": "Google Sheets account"
}
},
"typeVersion": 4.6
},
{
"id": "dd5a14f7-7321-4dfd-9f08-1f35efba1a07",
"name": "入力リマップ",
"type": "n8n-nodes-base.set",
"position": [
-1820,
100
],
"parameters": {
"options": {},
"assignments": {
"assignments": [
{
"id": "00924b90-278f-49f5-80f2-c297df0fcc97",
"name": "chatInput",
"type": "string",
"value": "={{ $json.input }}"
}
]
}
},
"typeVersion": 3.4
},
{
"id": "78d87631-7d52-4fbb-9313-f47263bd101f",
"name": "評価",
"type": "n8n-nodes-base.evaluation",
"position": [
-1180,
200
],
"parameters": {
"operation": "checkIfEvaluating"
},
"typeVersion": 4.6
},
{
"id": "1d8b12ba-ec63-4f92-a4dc-1885cbcd4d95",
"name": "入力フィールド設定",
"type": "n8n-nodes-base.set",
"position": [
-960,
100
],
"parameters": {
"options": {},
"assignments": {
"assignments": [
{
"id": "d58952c1-d346-4fbf-881e-d5c04b6781a5",
"name": "question",
"type": "string",
"value": "={{ $('When fetching a dataset row').first().json.input }}"
},
{
"id": "0f10a3d0-cf6e-4715-9ded-2cee54aa62ec",
"name": "answer",
"type": "string",
"value": "={{ $json.output }}"
},
{
"id": "edbe42ed-36a7-438a-989f-900673e61d0f",
"name": "groundTruth",
"type": "array",
"value": "={{ $('When fetching a dataset row').first().json['ground truth'].split('\\n') }}"
}
]
}
},
"typeVersion": 3.4
},
{
"id": "518cecea-9bcc-47dd-83b1-73ef0880e808",
"name": "操作なし（何もしない）",
"type": "n8n-nodes-base.noOp",
"position": [
-960,
300
],
"parameters": {},
"typeVersion": 1
},
{
"id": "c1a9cf2f-8e5d-4db9-a29b-b426e1932ab3",
"name": "AIエージェント",
"type": "@n8n/n8n-nodes-langchain.agent",
"position": [
-1600,
200
],
"parameters": {
"options": {}
},
"typeVersion": 2
},
{
"id": "6c5602d1-59c0-479e-b9aa-fb8ce037d4b1",
"name": "チャットメッセージ受信時",
"type": "@n8n/n8n-nodes-langchain.chatTrigger",
"position": [
-1820,
300
],
"webhookId": "ba1fadeb-b566-469a-97b3-3159a99f1805",
"parameters": {
"options": {}
},
"typeVersion": 1.1
},
{
"id": "6fb01095-7b93-41dd-8d7c-364cb685ddbb",
"name": "出力更新",
"type": "n8n-nodes-base.evaluation",
"position": [
520,
300
],
"parameters": {
"outputs": {
"values": [
{
"outputName": "output",
"outputValue": "={{ $('Set Input Fields').item.json.answer }}"
},
{
"outputName": "score",
"outputValue": "={{ $json.score }}"
}
]
},
"sheetName": {
"__rl": true,
"mode": "list",
"value": 1176192270,
"cachedResultUrl": "https://docs.google.com/spreadsheets/d/1YOnu2JJjlxd787AuYcg-wKbkjyjyZFgASYVV0jsij5Y/edit#gid=1176192270",
"cachedResultName": "Similarity"
},
"documentId": {
"__rl": true,
"mode": "list",
"value": "1YOnu2JJjlxd787AuYcg-wKbkjyjyZFgASYVV0jsij5Y",
"cachedResultUrl": "https://docs.google.com/spreadsheets/d/1YOnu2JJjlxd787AuYcg-wKbkjyjyZFgASYVV0jsij5Y/edit?usp=drivesdk",
"cachedResultName": "96. Evaluations Test"
}
},
"credentials": {
"googleSheetsOAuth2Api": {
"id": "XHvC7jIRR8A2TlUl",
"name": "Google Sheets account"
}
},
"typeVersion": 4.6
},
{
"id": "18e8dbb6-df98-4d36-9686-9ec481a37754",
"name": "メトリクス更新",
"type": "n8n-nodes-base.evaluation",
"position": [
700,
300
],
"parameters": {
"metrics": {
"assignments": [
{
"id": "1fd7759c-f4ef-4eda-87ad-9d9563b63e99",
"name": "score",
"type": "number",
"value": "={{ $json.score }}"
}
]
},
"operation": "setMetrics"
},
"typeVersion": 4.6
},
{
"id": "a685c9e7-0b21-4873-a502-5826a916e31f",
"name": "付箋1",
"type": "n8n-nodes-base.stickyNote",
"position": [
-2120,
-120
],
"parameters": {
"color": 7,
"width": 840,
"height": 720,
"content": "## 1. Setup Your AI Workflow to Use Evaluations\n[Learn more about the Evaluations Trigger](https://docs.n8n.io/integrations/builtin/?utm_source=n8n_app&utm_medium=node_settings_modal-credential_link&utm_campaign=n8n-nodes-base.evaluationTrigger)\n\nThe Evaluations Trigger is a separate execution which does not affect your production workflow in any way. It is manually triggered and automatically pulled datasets from the assigned Google Sheet."
},
"typeVersion": 1
},
{
"id": "2e57d911-8fef-456d-8e5c-3849350553e1",
"name": "付箋",
"type": "n8n-nodes-base.stickyNote",
"position": [
-1240,
-120
],
"parameters": {
"color": 7,
"width": 1700,
"height": 720,
"content": "## 2. Answer Similarity: How similar is the AI response to the Ground Truth?\n[Learn more about the Evaluations Trigger](https://docs.n8n.io/integrations/builtin/?utm_source=n8n_app&utm_medium=node_settings_modal-credential_link&utm_campaign=n8n-nodes-base.evaluationTrigger)\n\nFor this evaluation, we want to do a simple check to see how similar the AI's response is to the ground truth.\nThis test is most effective for closed-ended questions where there can be little to no deviation in the answers. "
},
"typeVersion": 1
},
{
"id": "9962bbf8-01bc-463c-ad3b-499618c5b1f9",
"name": "付箋3",
"type": "n8n-nodes-base.stickyNote",
"position": [
-2580,
-120
],
"parameters": {
"width": 420,
"height": 720,
"content": "## Try It Out!\n### This n8n template demonstrates how to calculate the evaluation metric \"Similarity\" which in this scenario, measures the consistency of the agent.\n\nThe scoring approach is adapted from [https://github.com/explodinggradients/ragas/blob/main/ragas/src/ragas/metrics/_answer_similarity.py](https://github.com/explodinggradients/ragas/blob/main/ragas/src/ragas/metrics/_answer_similarity.py)\n\n### How it works\n* This evaluation works best where questions are close-ended or about facts where the answer can have little to no deviation.\n* For our scoring, we generate embeddings for both the AI's response and ground truth and calculate the cosine similarity between them.\n* A high score indicates LLM consistency with expected results whereas a low score could signal model hallucination.\n\n### Requirements\n* n8n version 1.94+\n* Check out this Google Sheet for a sample data [https://docs.google.com/spreadsheets/d/1YOnu2JJjlxd787AuYcg-wKbkjyjyZFgASYVV0jsij5Y/edit?usp=sharing](https://docs.google.com/spreadsheets/d/1YOnu2JJjlxd787AuYcg-wKbkjyjyZFgASYVV0jsij5Y/edit?usp=sharing)\n\n\n### Need Help?\nJoin the [Discord](https://discord.com/invite/XPKeKXeB7d) or ask in the [Forum](https://community.n8n.io/)!\n\nHappy Hacking!"
},
"typeVersion": 1
},
{
"id": "47699677-545f-4eca-8b46-89c6862a56f1",
"name": "埋め込み取得",
"type": "n8n-nodes-base.httpRequest",
"position": [
-520,
300
],
"parameters": {
"url": "https://api.openai.com/v1/embeddings",
"method": "POST",
"options": {},
"sendBody": true,
"authentication": "predefinedCredentialType",
"bodyParameters": {
"parameters": [
{
"name": "input",
"value": "={{ $json.groundTruth }}"
},
{
"name": "model",
"value": "text-embedding-3-small"
},
{
"name": "encoding_format",
"value": "float"
}
]
},
"nodeCredentialType": "openAiApi"
},
"credentials": {
"openAiApi": {
"id": "8gccIjcuf3gvaoEr",
"name": "OpenAi account"
}
},
"typeVersion": 4.2
},
{
"id": "1a022046-80f2-4df3-b075-be8e9294b951",
"name": "GroundTruthをアイテムに変換",
"type": "n8n-nodes-base.splitOut",
"position": [
-720,
300
],
"parameters": {
"options": {},
"fieldToSplitOut": "groundTruth"
},
"typeVersion": 1
},
{
"id": "0cea3a79-23ab-4d0f-9c6c-2522af6a506f",
"name": "埋め込み取得1",
"type": "n8n-nodes-base.httpRequest",
"position": [
-520,
100
],
"parameters": {
"url": "https://api.openai.com/v1/embeddings",
"method": "POST",
"options": {},
"sendBody": true,
"authentication": "predefinedCredentialType",
"bodyParameters": {
"parameters": [
{
"name": "input",
"value": "={{ $json.answer }}"
},
{
"name": "model",
"value": "text-embedding-3-small"
},
{
"name": "encoding_format",
"value": "float"
}
]
},
"nodeCredentialType": "openAiApi"
},
"credentials": {
"openAiApi": {
"id": "8gccIjcuf3gvaoEr",
"name": "OpenAi account"
}
},
"typeVersion": 4.2
},
{
"id": "1b5740cc-4b79-4690-88ae-fa31c5a1c16c",
"name": "集計",
"type": "n8n-nodes-base.aggregate",
"position": [
-120,
300
],
"parameters": {
"options": {},
"aggregate": "aggregateAllItemData",
"destinationFieldName": "groundTruth"
},
"typeVersion": 1
},
{
"id": "d32f979b-7d96-4730-8324-de0d34ed659a",
"name": "埋め込みリマップ",
"type": "n8n-nodes-base.set",
"position": [
-320,
100
],
"parameters": {
"options": {},
"assignments": {
"assignments": [
{
"id": "3db07c98-f926-46e0-85f2-ed1eb137f842",
"name": "answer",
"type": "array",
"value": "={{ $json.data[0].embedding }}"
}
]
}
},
"typeVersion": 3.4
},
{
"id": "188864fb-3d6a-4c01-9d23-73c731de023d",
"name": "埋め込みリマップ1",
"type": "n8n-nodes-base.set",
"position": [
-320,
300
],
"parameters": {
"options": {},
"assignments": {
"assignments": [
{
"id": "3db07c98-f926-46e0-85f2-ed1eb137f842",
"name": "data",
"type": "array",
"value": "={{ $json.data[0].embedding }}"
}
]
}
},
"typeVersion": 3.4
},
{
"id": "5cc6c1c7-240b-4129-b57a-e7d0ef2e72e6",
"name": "埋め込み結果作成",
"type": "n8n-nodes-base.merge",
"position": [
80,
300
],
"parameters": {
"mode": "combine",
"options": {},
"combineBy": "combineByPosition"
},
"typeVersion": 3.1
},
{
"id": "7aa037d2-31a6-4efe-81b1-26a95c4c6d38",
"name": "類似度スコア計算",
"type": "n8n-nodes-base.code",
"position": [
280,
300
],
"parameters": {
"jsCode": "const { answer, groundTruth = [] } = $input.item.json;\n\nconst scores = await Promise.all(groundTruth.map(truth =>\n cosineSimilarity(answer, truth.data)\n));\n\nconst averageScore = scores.reduce((acc,i) => acc + i, 0) / scores.length;\n\nreturn { json: { score: averageScore } };\n\nfunction cosineSimilarity(a, b) { \n let dotProduct = normA = normB = 0;\n for (let i = 0; i < a.length; i++) {\n dotProduct += a[i] * b[i];\n normA += a[i] ** 2;\n normB += b[i] ** 2;\n }\n return dotProduct / (Math.sqrt(normA) * Math.sqrt(normB));\n}"
},
"typeVersion": 2
}
],
"pinData": {},
"connections": {
"c1a9cf2f-8e5d-4db9-a29b-b426e1932ab3": {
"main": [
[
{
"node": "78d87631-7d52-4fbb-9313-f47263bd101f",
"type": "main",
"index": 0
}
]
]
},
"1b5740cc-4b79-4690-88ae-fa31c5a1c16c": {
"main": [
[
{
"node": "5cc6c1c7-240b-4129-b57a-e7d0ef2e72e6",
"type": "main",
"index": 1
}
]
]
},
"78d87631-7d52-4fbb-9313-f47263bd101f": {
"main": [
[
{
"node": "1d8b12ba-ec63-4f92-a4dc-1885cbcd4d95",
"type": "main",
"index": 0
}
],
[
{
"node": "518cecea-9bcc-47dd-83b1-73ef0880e808",
"type": "main",
"index": 0
}
]
]
},
"dd5a14f7-7321-4dfd-9f08-1f35efba1a07": {
"main": [
[
{
"node": "c1a9cf2f-8e5d-4db9-a29b-b426e1932ab3",
"type": "main",
"index": 0
}
]
]
},
"6fb01095-7b93-41dd-8d7c-364cb685ddbb": {
"main": [
[
{
"node": "18e8dbb6-df98-4d36-9686-9ec481a37754",
"type": "main",
"index": 0
}
]
]
},
"47699677-545f-4eca-8b46-89c6862a56f1": {
"main": [
[
{
"node": "188864fb-3d6a-4c01-9d23-73c731de023d",
"type": "main",
"index": 0
}
]
]
},
"0cea3a79-23ab-4d0f-9c6c-2522af6a506f": {
"main": [
[
{
"node": "d32f979b-7d96-4730-8324-de0d34ed659a",
"type": "main",
"index": 0
}
]
]
},
"d32f979b-7d96-4730-8324-de0d34ed659a": {
"main": [
[
{
"node": "5cc6c1c7-240b-4129-b57a-e7d0ef2e72e6",
"type": "main",
"index": 0
}
]
]
},
"1d8b12ba-ec63-4f92-a4dc-1885cbcd4d95": {
"main": [
[
{
"node": "0cea3a79-23ab-4d0f-9c6c-2522af6a506f",
"type": "main",
"index": 0
},
{
"node": "1a022046-80f2-4df3-b075-be8e9294b951",
"type": "main",
"index": 0
}
]
]
},
"188864fb-3d6a-4c01-9d23-73c731de023d": {
"main": [
[
{
"node": "1b5740cc-4b79-4690-88ae-fa31c5a1c16c",
"type": "main",
"index": 0
}
]
]
},
"c5365531-2d66-4e67-9db2-37059e9e97a3": {
"ai_languageModel": [
[
{
"node": "c1a9cf2f-8e5d-4db9-a29b-b426e1932ab3",
"type": "ai_languageModel",
"index": 0
}
]
]
},
"1a022046-80f2-4df3-b075-be8e9294b951": {
"main": [
[
{
"node": "47699677-545f-4eca-8b46-89c6862a56f1",
"type": "main",
"index": 0
}
]
]
},
"5cc6c1c7-240b-4129-b57a-e7d0ef2e72e6": {
"main": [
[
{
"node": "7aa037d2-31a6-4efe-81b1-26a95c4c6d38",
"type": "main",
"index": 0
}
]
]
},
"7aa037d2-31a6-4efe-81b1-26a95c4c6d38": {
"main": [
[
{
"node": "6fb01095-7b93-41dd-8d7c-364cb685ddbb",
"type": "main",
"index": 0
}
]
]
},
"6c5602d1-59c0-479e-b9aa-fb8ce037d4b1": {
"main": [
[
{
"node": "c1a9cf2f-8e5d-4db9-a29b-b426e1932ab3",
"type": "main",
"index": 0
}
]
]
},
"3881b831-b575-4b2f-8ebb-ac956a714f10": {
"main": [
[
{
"node": "dd5a14f7-7321-4dfd-9f08-1f35efba1a07",
"type": "main",
"index": 0
}
]
]
}
}
}よくある質問
このワークフローの使い方は？
上記のJSON設定コードをコピーし、n8nインスタンスで新しいワークフローを作成して「JSONからインポート」を選択、設定を貼り付けて認証情報を必要に応じて変更してください。
このワークフローはどんな場面に適していますか？
上級 - エンジニアリング, 人工知能
有料ですか？
このワークフローは完全無料です。ただし、ワークフローで使用するサードパーティサービス（OpenAI APIなど）は別途料金が発生する場合があります。
関連ワークフロー
AIエージェントの応答正確性をOpenAIとRAGASメソッドで評価
OpenAIとRAGASメソッドを使用してAIエージェントの応答の正確性を評価する
Set
Code
Merge
+
Set
Code
Merge
27 ノードJimleuk
エンジニアリング
AIエージェントの応答関連性をOpenAIとコサイン類似度で評価
OpenAIとコサイン類似度を使用してAIエージェントの応答の関連性を評価する
Set
Code
Evaluation
+
Set
Code
Evaluation
20 ノードJimleuk
エンジニアリング
OpenAIを使ったRAG応答の精度評価：ドキュメント・ベースライン指標
OpenAIによるRAG応答の精度評価：ドキュメントベースライン指標
Set
Evaluation
Http Request
+
Set
Evaluation
Http Request
25 ノードJimleuk
エンジニアリング
Voyage-Context-3 エンベディングと MongoDB Atlas に基づくドキュメントQ&Aシステム
Voyage-Context-3埋め込みモデルおよびMongoDB Atlasを基盤としたドキュメントQ&Aシステム
Set
Code
Wait
+
Set
Code
Wait
53 ノードJimleuk
エンジニアリング
n8n、Apify、OpenAI o3 を使用したセルフホスト型 AI ディープリサーチエージェント
n8n、Apify、OpenAI o3を使用したセルフホスト型AI深度リサーチエージェント
If
Set
Code
+
If
Set
Code
87 ノードJimleuk
人工知能
Excel、Outlook、AIを使用した日中のニュースサマリーサービス
Excel、Outlook、AIを使った毎日のニュースブリーフィングサービス
If
Set
Html
+
If
Set
Html
33 ノードJimleuk
人工知能
ワークフロー情報
難易度
上級
ノード数21
カテゴリー2
ノードタイプ13
作成者
Jimleuk
@jimleukFreelance consultant based in the UK specialising in AI-powered automations. I work with select clients tackling their most challenging projects. For business enquiries, send me an email at hello@jimle.uk LinkedIn: https://www.linkedin.com/in/jimleuk/ X/Twitter: https://x.com/jimle_uk
外部リンク
n8n.ioで表示 →
このワークフローを共有