GPT-4、ウィキペディア、ベクター検索エンジンを使用して包括のなエンティティプロファイルを構築
上級
これはDocument Extraction, AI RAG分野の自動化ワークフローで、33個のノードを含みます。主にIf, Set, Merge, ManualTrigger, Agentなどのノードを使用。 GPT-4、ウィキペディア、ベクトルデータベースを使って、コンテンツの包括のエンティティアーカイブを構築
前提条件
- •OpenAI API Key
- •Qdrantサーバー接続情報
使用ノード (33)
カテゴリー
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": "6fb6c9a50faeb88c76c44b2fdb3a06e4272886afd55ba8d161b7b55a4373c282",
"templateCredsSetupCompleted": true
},
"nodes": [
{
"id": "8b99d8c3-eb24-436a-8f64-2eca763ca38c",
"name": "「ワークフロー実行」クリック時",
"type": "n8n-nodes-base.manualTrigger",
"position": [
-768,
876
],
"parameters": {},
"typeVersion": 1
},
{
"id": "bd998532-090f-4ec3-b3a8-c942dbd4d77c",
"name": "他ワークフローからの実行時",
"type": "n8n-nodes-base.executeWorkflowTrigger",
"position": [
-768,
684
],
"parameters": {
"workflowInputs": {
"values": [
{
"name": "entity"
},
{
"name": "topic"
},
{
"name": "audience"
},
{
"name": "purpose"
},
{
"name": "execution_id",
"type": "number"
}
]
}
},
"typeVersion": 1.1
},
{
"id": "12f3159a-6df5-4603-a9b0-5c742319e6da",
"name": "OpenAI Chat Model2",
"type": "@n8n/n8n-nodes-langchain.lmChatOpenAi",
"position": [
256,
1004
],
"parameters": {
"model": {
"__rl": true,
"mode": "list",
"value": "o4-mini",
"cachedResultName": "o4-mini"
},
"options": {}
},
"credentials": {
"openAiApi": {
"id": "Y8ZxcY3KmZ6Zqrd2",
"name": "OpenAi account"
}
},
"typeVersion": 1.2
},
{
"id": "e5ece6f2-9476-459d-a153-9c27aa0d6a9b",
"name": "OpenAI Chat Model4",
"type": "@n8n/n8n-nodes-langchain.lmChatOpenAi",
"position": [
1368,
880
],
"parameters": {
"model": {
"__rl": true,
"mode": "list",
"value": "o4-mini",
"cachedResultName": "o4-mini"
},
"options": {}
},
"credentials": {
"openAiApi": {
"id": "Y8ZxcY3KmZ6Zqrd2",
"name": "OpenAi account"
}
},
"typeVersion": 1.2
},
{
"id": "45afcd19-cd43-47b0-b78b-8ecfda956359",
"name": "質問回答済み",
"type": "n8n-nodes-base.if",
"position": [
1136,
780
],
"parameters": {
"options": {},
"conditions": {
"options": {
"version": 2,
"leftValue": "",
"caseSensitive": true,
"typeValidation": "strict"
},
"combinator": "and",
"conditions": [
{
"id": "6197485c-b2ec-4904-87a0-f52185e1305d",
"operator": {
"type": "number",
"operation": "gte"
},
"leftValue": "={{ $json.output.confidence }}",
"rightValue": 0
}
]
}
},
"typeVersion": 2.2
},
{
"id": "9ce6eceb-d170-4ee4-897f-426f9adbf48b",
"name": "Wikipedia",
"type": "@n8n/n8n-nodes-langchain.toolWikipedia",
"position": [
384,
1004
],
"parameters": {},
"typeVersion": 1
},
{
"id": "31f278bf-12b5-41ee-b200-a4f8e56796c9",
"name": "文字テキスト分割器",
"type": "@n8n/n8n-nodes-langchain.textSplitterCharacterTextSplitter",
"position": [
2944,
912
],
"parameters": {
"chunkSize": 10000
},
"typeVersion": 1
},
{
"id": "0218c8e3-21ec-4584-b51c-ac47b2f85364",
"name": "デフォルトデータローダー",
"type": "@n8n/n8n-nodes-langchain.documentDefaultDataLoader",
"position": [
2864,
704
],
"parameters": {
"options": {
"metadata": {
"metadataValues": [
{
"name": "entity_name",
"value": "={{ $('Merge').item.json.output.entity_name.toLowerCase() }}"
},
{
"name": "type",
"value": "={{ $('Merge').item.json.output.type }}"
},
{
"name": "definition",
"value": "={{ $('Merge').item.json.output.definition }}"
},
{
"name": "category",
"value": "={{ $('Merge').item.json.output.category }}"
},
{
"name": "relevance",
"value": "={{ $('Merge').item.json.output.relevance }}"
},
{
"name": "alternative_names",
"value": "={{ $('Merge').item.json.output.alternative_names }}"
},
{
"name": "example",
"value": "={{ $('Merge').item.json.output.example }}"
},
{
"name": "reference_link",
"value": "={{ $('Merge').item.json.output.reference_link }}"
},
{
"name": "misconceptions",
"value": "={{ $('Merge').item.json.output.misconceptions }}"
},
{
"name": "=related_entities",
"value": "={{ $('Merge').item.json.output.related_entities }}"
},
{
"name": "ai_model_used",
"value": "={{ $('Merge').item.json.output.ai_model_used }}"
},
{
"name": "confidence",
"value": "={{ $('Merge').item.json.output.confidence }}"
}
]
}
},
"jsonData": "={{ $('Merge').item.json.output.entity_name.toLowerCase() }}",
"jsonMode": "expressionData",
"textSplittingMode": "custom"
},
"typeVersion": 1.1
},
{
"id": "c95360a1-81f7-4bf9-81c9-bcac1d918ad1",
"name": "マージ1",
"type": "n8n-nodes-base.merge",
"position": [
3232,
832
],
"parameters": {
"mode": "chooseBranch"
},
"typeVersion": 3.2
},
{
"id": "43e0fc70-b7dc-4148-b3af-02af36ef4775",
"name": "リサーチャーインターネット",
"type": "@n8n/n8n-nodes-langchain.toolWorkflow",
"position": [
512,
1004
],
"parameters": {
"workflowId": {
"__rl": true,
"mode": "list",
"value": "hRo7kt0ghoXmEv4G",
"cachedResultName": "Researcher: Internet"
},
"workflowInputs": {
"value": {},
"schema": [
{
"id": "query",
"type": "string",
"display": true,
"removed": false,
"required": false,
"displayName": "query",
"defaultMatch": false,
"canBeUsedToMatch": true
}
],
"mappingMode": "defineBelow",
"matchingColumns": [
"query"
],
"attemptToConvertTypes": false,
"convertFieldsToString": false
}
},
"typeVersion": 2.2
},
{
"id": "013f6548-399d-43ad-a738-f280c0f3621a",
"name": "マージ",
"type": "n8n-nodes-base.merge",
"position": [
1936,
780
],
"parameters": {
"mode": "chooseBranch",
"useDataOfInput": 2
},
"typeVersion": 3.2
},
{
"id": "d3c62700-d7d3-4152-a1eb-8fd4d5b7278d",
"name": "エンティティ検索",
"type": "@n8n/n8n-nodes-langchain.vectorStoreQdrant",
"onError": "continueRegularOutput",
"position": [
-320,
780
],
"parameters": {
"mode": "load",
"prompt": "={{ $json.entity.toLowerCase() }}",
"options": {
"searchFilterJson": "={\n \"must\": [\n {\n \"key\": \"metadata.entity_name\",\n \"match\": {\n \"value\": \"{{ $json.entity.toLowerCase() }}\"\n }\n }\n ]\n}"
},
"qdrantCollection": {
"__rl": true,
"mode": "id",
"value": "={{ $json.collection_name }}"
}
},
"credentials": {
"qdrantApi": {
"id": "1rlzgQcdRysKi0oS",
"name": "QdrantApi account"
}
},
"typeVersion": 1.3,
"alwaysOutputData": true
},
{
"id": "db69e156-1c18-48a3-9178-9f8341a73bfc",
"name": "エンティティ検索埋め込み",
"type": "@n8n/n8n-nodes-langchain.embeddingsOllama",
"position": [
-248,
1004
],
"parameters": {
"model": "nomic-embed-text:latest"
},
"credentials": {
"ollamaApi": {
"id": "u58LNTOTwwLnJzt9",
"name": "Ollama account"
}
},
"typeVersion": 1
},
{
"id": "a28aeee2-3422-4455-a39a-84ebc0d57efd",
"name": "エンティティ検索成功",
"type": "n8n-nodes-base.if",
"onError": "continueRegularOutput",
"position": [
32,
780
],
"parameters": {
"options": {},
"conditions": {
"options": {
"version": 2,
"leftValue": "",
"caseSensitive": true,
"typeValidation": "strict"
},
"combinator": "and",
"conditions": [
{
"id": "d1f73505-50c0-4762-8f5a-3af4f8511a51",
"operator": {
"type": "object",
"operation": "exists",
"singleValue": true
},
"leftValue": "={{ $json.document }}",
"rightValue": "={{ $json.document }}"
}
]
}
},
"typeVersion": 2.2,
"alwaysOutputData": false
},
{
"id": "f70fac9a-d478-42a0-a14b-5ff6cf864c57",
"name": "エンティティ保存",
"type": "@n8n/n8n-nodes-langchain.vectorStoreQdrant",
"onError": "continueRegularOutput",
"position": [
2768,
480
],
"parameters": {
"mode": "insert",
"options": {},
"qdrantCollection": {
"__rl": true,
"mode": "list",
"value": "entities",
"cachedResultName": "entities"
},
"embeddingBatchSize": 1
},
"credentials": {
"qdrantApi": {
"id": "1rlzgQcdRysKi0oS",
"name": "QdrantApi account"
}
},
"typeVersion": 1.3
},
{
"id": "e1fd4754-75cb-4d44-bc7b-01d1eddfdb75",
"name": "エンティティ検索2",
"type": "@n8n/n8n-nodes-langchain.vectorStoreQdrant",
"onError": "continueRegularOutput",
"position": [
2160,
656
],
"parameters": {
"mode": "load",
"prompt": "={{ $json.output.entity_name.toLowerCase() }}",
"options": {
"searchFilterJson": "{\n \"must\": [\n {\n \"key\": \"metadata.entity_name\",\n \"match\": {\n \"value\": \"{{ $json.output.entity_name.toLowerCase() }}\"\n }\n }\n ]\n}"
},
"qdrantCollection": {
"__rl": true,
"mode": "id",
"value": "={{ $('Set Node').item.json.collection_name }}"
}
},
"credentials": {
"qdrantApi": {
"id": "1rlzgQcdRysKi0oS",
"name": "QdrantApi account"
}
},
"typeVersion": 1.3,
"alwaysOutputData": true
},
{
"id": "5a6e8cd3-c68d-4f24-9f3d-d55a8a0a9267",
"name": "エンティティ定義済み",
"type": "n8n-nodes-base.if",
"position": [
1712,
656
],
"parameters": {
"options": {},
"conditions": {
"options": {
"version": 2,
"leftValue": "",
"caseSensitive": true,
"typeValidation": "strict"
},
"combinator": "and",
"conditions": [
{
"id": "6197485c-b2ec-4904-87a0-f52185e1305d",
"operator": {
"type": "boolean",
"operation": "true",
"singleValue": true
},
"leftValue": "={{ $json.output[\"is_complete\"] }}",
"rightValue": ""
}
]
}
},
"typeVersion": 2.2
},
{
"id": "85c800e2-a0da-42e9-aaf7-b2665aeb0969",
"name": "エンティティ検証",
"type": "@n8n/n8n-nodes-langchain.chainLlm",
"maxTries": 5,
"position": [
1360,
656
],
"parameters": {
"text": "=entity_name: {{ $json.output.entity_name }}\ntype: {{ $json.output.type }}\ndefinition: {{ $json.output.definition }}\ncategory: {{ $json.output.category }}\nrelevance: {{ $json.output.relevance }}\nalternative_names: {{ $json.output.alternative_names }}\nexample: {{ $json.output.example }}\nreference_link: {{ $json.output.reference_link }}\nrelated entities: {{ $json.output.related_entities }}\nmisconceptions: {{ $json.output.misconceptions }}",
"batching": {},
"messages": {
"messageValues": [
{
"message": "=You are an entity quality evaluator.\n\nYour job is to judge if an entity record is complete—that is, if it provides all key fields and gives a clear, direct, and accurate description of the entity for the intended business audience.\n\nScoring Rules:\n\t•\tMark as is_complete: true ONLY if all main fields are filled with relevant, plain-English information (not empty or obviously generic), and if the definition, type, category, and relevance are clear and correct for a non-technical business user.\n\t•\tMark as is_complete: false if any major field (definition, type, category, relevance, example, reference link) is missing, vague, off-topic, generic, or just restates the entity name.\n\t•\tAccept an empty array for alternative_names, misconceptions, or related_entities, but do not mark as complete if these fields are simply omitted (missing from the object).\n\t•\tIf you find factual mistakes, off-topic content, or AI hallucination, mark as incomplete and explain.\n\t•\tIf the entity is not explained in a way a typical business user could understand, mark as incomplete.\n\nResponse Format:\nReply with a valid JSON object, and nothing else:\n\n{ \"is_complete\": true/false, \"reason\": \"short explanation\" }\n\nDo not add any extra text, comments, or formatting outside the JSON."
}
]
},
"promptType": "define",
"hasOutputParser": true
},
"retryOnFail": true,
"typeVersion": 1.7,
"waitBetweenTries": 500
},
{
"id": "dcb79c8f-07ab-4690-ac10-c2aa7d0f08d9",
"name": "エンティティ存在確認",
"type": "n8n-nodes-base.if",
"position": [
2512,
656
],
"parameters": {
"options": {},
"conditions": {
"options": {
"version": 2,
"leftValue": "",
"caseSensitive": true,
"typeValidation": "strict"
},
"combinator": "and",
"conditions": [
{
"id": "d1f73505-50c0-4762-8f5a-3af4f8511a51",
"operator": {
"type": "object",
"operation": "notEmpty",
"singleValue": true
},
"leftValue": "={{ $json.document }}",
"rightValue": "={{ $json.document }}"
},
{
"id": "4cfb2390-5c31-4cae-bbe0-15a9310a5d5a",
"operator": {
"name": "filter.operator.equals",
"type": "string",
"operation": "equals"
},
"leftValue": "",
"rightValue": ""
}
]
}
},
"typeVersion": 2.2
},
{
"id": "944e691a-6333-4382-8f96-2e85dd199c54",
"name": "エンティティリサーチャー",
"type": "@n8n/n8n-nodes-langchain.agent",
"position": [
440,
780
],
"parameters": {
"text": "=Entity Research Request:\n\nResearch and deliver a writer-ready entity profile for the following:\n\t•\tEntity: {{ $('Set Node').item.json.entity }}\n\t•\tContext/Topic: {{ $('Set Node').item.json.topic }}\n\t•\tAudience: {{ $('Set Node').item.json.audience }}\n\nInstructions:\nBuild a practical brief for our writing team. If you use any research tools (Wikipedia or Internet Researcher), always use only the entity name as your search/query term—not the context or audience.\nReturn only the required JSON object as output.\n",
"options": {
"systemMessage": "=You are an expert entity researcher.\nYour job is to deliver a complete, practical entity brief—ready for a writing team—using this information-gathering protocol:\n\nTool Use Protocol:\n\t1.\tStart with the vector database for all available details on the entity.\n\t2.\tIf needed, use Wikipedia to fill in missing details or clarify points.\nWhen querying Wikipedia, use only the entity name as your search term.\n\t3.\tIf essential information is still missing, use the Internet Researcher tool (up to two times).\nWhen using Internet Researcher, always use only the entity name as your query.\nIf after two queries information is still unavailable, leave those fields blank or as empty arrays.\n\t4.\tNever include any tool usage details, research path, or confidence scores in your output.\nOnly return the required JSON object.\n\nResearch priorities:\n\t•\tWrite in plain English for the target audience.\n\t•\tProvide real-world examples.\n\t•\tAddress common misconceptions when relevant.\n\t•\tBe concise but thorough—no padding.\n\nCritical Reminders:\n\t•\tIf information is unavailable after all tool steps, use empty arrays/strings.\n\t•\tNever include commentary, tool usage, or confidence scores.\n\t•\tDo not add markdown or extra explanation—JSON only.\n",
"returnIntermediateSteps": false
},
"promptType": "define",
"hasOutputParser": true
},
"retryOnFail": true,
"typeVersion": 2,
"waitBetweenTries": 5000
},
{
"id": "7e28cd46-869d-4433-acdc-57bc10da7935",
"name": "付箋",
"type": "n8n-nodes-base.stickyNote",
"position": [
-1760,
144
],
"parameters": {
"width": 848,
"height": 1056,
"content": "### This n8n template demonstrates how to build an intelligent entity research system that automatically discovers, researches, and creates comprehensive profiles for business entities, concepts, and terms.\nPerfect for content teams, business analysts, compliance officers, or anyone who needs consistent, authoritative definitions and explanations of industry terms, products, technologies, or concepts!\n\n### How it works\n* **Smart Entity Discovery**: Before researching, the workflow checks your existing knowledge base to avoid duplicate work and build upon existing information.\n* **Multi-Source Research**: Uses a sophisticated AI agent that intelligently combines your vector database, Wikipedia, and live web research to gather comprehensive entity information.\n* **Structured Entity Profiles**: Creates consistent, business-ready entity profiles with definitions, categories, examples, misconceptions, and related entities - perfect for glossaries, training materials, or content reference.\n* **Quality Validation**: AI-powered validation ensures all entity profiles are complete, accurate, and suitable for business use before storage.\n* **Incremental Knowledge Building**: Each researched entity gets stored in your vector database, creating a growing knowledge base that improves over time.\n* **Duplicate Prevention**: Multi-stage checking prevents duplicate entities and unnecessary re-research, saving time and API costs.\n\n### How to use\n* **Simple Start**: Use the manual trigger with any entity like \"OAuth 2.0\", \"GDPR\", \"Machine Learning\", or \"API Gateway\"\n* **Workflow Integration**: Replace the manual trigger with form submissions, content management systems, or automated content pipelines\n* **Topic-Specific Research**: Provide topic and audience context to get tailored explanations (e.g., \"blockchain\" for \"IT managers\" vs \"general audience\")\n* **Batch Processing**: Process multiple entities at once for building comprehensive glossaries or knowledge bases\n\n### Requirements\n* **OpenAI API** - For o4-mini (entity research and validation)\n* **Anthropic API** - For Claude Sonnet 4 (quality validation and completeness checking)\n* **Qdrant Vector Database** - For entity storage and duplicate detection\n* **Ollama** - For local embeddings (nomic-embed-text model)\n* **Wikipedia Access** - For foundational research (built-in to n8n)\n* **Internet Research Tool** - Optional connection to web research workflow for current information\n\n### Configuration Notes\n⚠️ **Database Collections**: Ensure Qdrant collection \"entities\" exists and is properly configured\n⚠️ **Environment**: Update localhost URLs to your actual Qdrant instance address\n⚠️ **API Limits**: Monitor usage as each entity research can use 2-5 API calls depending on complexity\n\n### Perfect for\n* **Content Teams**: Building consistent entity definitions for articles and documentation\n* **Business Analysts**: Creating standardized definitions for business processes and technologies\n* **Training Departments**: Developing comprehensive glossaries for employee education\n* **Compliance Teams**: Maintaining up-to-date definitions of regulatory terms and concepts\n* **Product Teams**: Creating clear explanations of features, technologies, and industry concepts\n* **Knowledge Management**: Building searchable databases of business-critical concepts\n"
},
"typeVersion": 1
},
{
"id": "688c507f-bc3b-42ee-9cbd-22f77031e3ca",
"name": "付箋1",
"type": "n8n-nodes-base.stickyNote",
"position": [
-448,
608
],
"parameters": {
"width": 528,
"height": 144,
"content": "## Smart Entity Lookup\nThis checks if the entity has already been researched before. If it finds existing information, it skips the research and uses what's already available. If nothing is found, the workflow continues to research the entity from scratch."
},
"typeVersion": 1
},
{
"id": "63efd98f-9b03-4fb9-94f4-66331f4e1a98",
"name": "付箋2",
"type": "n8n-nodes-base.stickyNote",
"position": [
384,
592
],
"parameters": {
"width": 592,
"height": 176,
"content": "## AI Research Agent\n\nThis is the core research engine that creates comprehensive entity profiles using multiple information sources. The AI agent automatically searches your knowledge database first, then uses Wikipedia and web research as needed to build complete entity definitions with examples, misconceptions, and related terms."
},
"typeVersion": 1
},
{
"id": "ac41348b-f548-42cc-92b9-ad958d3a5314",
"name": "付箋3",
"type": "n8n-nodes-base.stickyNote",
"position": [
1472,
448
],
"parameters": {
"width": 416,
"height": 176,
"content": "## Quality Control Validator\n\nChecks if the researched entity profile is complete and accurate before saving it. The AI validator ensures all validates fields are filled with meaningful information and that the definition is clear enough for business users to understand."
},
"typeVersion": 1
},
{
"id": "75815fc7-3dc6-4008-a604-03e080fc4740",
"name": "付箋4",
"type": "n8n-nodes-base.stickyNote",
"position": [
2336,
448
],
"parameters": {
"width": 400,
"height": 176,
"content": "## Final Duplicate Check\n\nThis performs a second search to see if the newly researched entity already exists in the database before saving it. It prevents creating duplicate entries when similar entities might have been found during the research process."
},
"typeVersion": 1
},
{
"id": "fdd41059-8c96-40b9-9cb0-3f30c0da35d1",
"name": "付箋5",
"type": "n8n-nodes-base.stickyNote",
"position": [
2816,
272
],
"parameters": {
"width": 400,
"height": 176,
"content": "## Entity Storage\n\nThis saves the validated entity profile to the knowledge database for future use. It packages all the entity information (definition, examples, related terms, etc.) and stores it in a searchable format so it can be found and reused later."
},
"typeVersion": 1
},
{
"id": "cd787fbd-d7f6-411a-bd98-af776eeb94b3",
"name": "設定ノード",
"type": "n8n-nodes-base.set",
"position": [
-544,
780
],
"parameters": {
"options": {},
"assignments": {
"assignments": [
{
"id": "5833b571-4261-4775-abd7-2f2fb9fbb015",
"name": "entity",
"type": "string",
"value": "={{ $json.entity }}"
},
{
"id": "21a70173-fb72-4c98-a9ac-566e5cc0e3be",
"name": "topic",
"type": "string",
"value": "={{ $json.topic }}"
},
{
"id": "eec9aea6-b681-4656-bd14-c68456f5c2d9",
"name": "audience",
"type": "string",
"value": "={{ $json.audience }}"
},
{
"id": "bd6b2273-a10e-42cc-a834-f141c360fbd6",
"name": "purpose",
"type": "string",
"value": "={{ $json.purpose }}"
},
{
"id": "05dae7d9-fff6-44a2-a2a2-cd3426c08021",
"name": "execution_id",
"type": "string",
"value": "={{ $json.execution_id }}"
},
{
"id": "ea6e7b26-e236-426b-879e-fb8c82d79809",
"name": "collection_name",
"type": "string",
"value": "entities"
}
]
}
},
"typeVersion": 3.4
},
{
"id": "0eb308dc-5756-488e-a941-c2c51b87e08b",
"name": "エンティティ検索1",
"type": "@n8n/n8n-nodes-langchain.vectorStoreQdrant",
"position": [
640,
1004
],
"parameters": {
"mode": "retrieve-as-tool",
"options": {},
"toolDescription": "Search for entity details",
"qdrantCollection": {
"__rl": true,
"mode": "id",
"value": "={{ $('Set Node').item.json.collection_name }}"
}
},
"credentials": {
"qdrantApi": {
"id": "1rlzgQcdRysKi0oS",
"name": "QdrantApi account"
}
},
"typeVersion": 1.3
},
{
"id": "f1d39f85-4e48-43c6-ab33-9c30517a5c3e",
"name": "エンティティ検索埋め込み1",
"type": "@n8n/n8n-nodes-langchain.embeddingsOllama",
"position": [
720,
1212
],
"parameters": {
"model": "nomic-embed-text:latest"
},
"credentials": {
"ollamaApi": {
"id": "u58LNTOTwwLnJzt9",
"name": "Ollama account"
}
},
"typeVersion": 1
},
{
"id": "0353855d-d641-4bec-97cc-c8e53be5d9b6",
"name": "エンティティパーサー",
"type": "@n8n/n8n-nodes-langchain.outputParserStructured",
"position": [
928,
1004
],
"parameters": {
"jsonSchemaExample": "{\n \"can_answer\": true,\n \"entity_name\": \"Gmail\",\n \"type\": \"Product\",\n \"definition\": \"Gmail is Google's cloud-based email service for personal and business use, offering web and mobile access.\",\n \"category\": \"Software\",\n \"relevance\": \"Gmail is widely used and is a frequent target for cybersecurity threats.\",\n \"alternative_names\": [\"Google Mail\"],\n \"example\": \"A user receives a phishing attempt in their Gmail inbox.\",\n \"reference_link\": \"https://mail.google.com/\",\n \"misconceptions\": [\"Gmail accounts cannot be compromised.\", \"Only paid Gmail accounts are secure.\"],\n \"related_entities\": [\"Google Workspace\", \"Two-Factor Authentication (2FA)\", \"Phishing Prevention\"],\n \"ai_model_used\": \"04-mini\",\n \"confidence\": 0.97\n}"
},
"typeVersion": 1.3
},
{
"id": "b9161937-611c-46d5-b372-1a6f0186b637",
"name": "検証パーサー",
"type": "@n8n/n8n-nodes-langchain.outputParserStructured",
"position": [
1496,
880
],
"parameters": {
"jsonSchemaExample": "{\n\t\"is_complete\": true,\n\t\"reason\": \"The answer clearly and directly states California’s location and borders, fully addressing the question.\"\n}"
},
"typeVersion": 1.3
},
{
"id": "a084c458-9089-41eb-8017-5523891e159f",
"name": "エンティティ検索埋め込み2",
"type": "@n8n/n8n-nodes-langchain.embeddingsOllama",
"position": [
2232,
880
],
"parameters": {
"model": "nomic-embed-text:latest"
},
"credentials": {
"ollamaApi": {
"id": "u58LNTOTwwLnJzt9",
"name": "Ollama account"
}
},
"typeVersion": 1
},
{
"id": "26684a88-c3d6-407f-af17-b54a25ca5da8",
"name": "エンティティ埋め込み",
"type": "@n8n/n8n-nodes-langchain.embeddingsOllama",
"position": [
2736,
704
],
"parameters": {
"model": "nomic-embed-text:latest"
},
"credentials": {
"ollamaApi": {
"id": "u58LNTOTwwLnJzt9",
"name": "Ollama account"
}
},
"typeVersion": 1
}
],
"pinData": {
"When clicking ‘Execute workflow’": [
{
"topic": "",
"entity": "Darth Vader",
"purpose": "",
"audience": "",
"execution_id": 1
}
]
},
"connections": {
"013f6548-399d-43ad-a738-f280c0f3621a": {
"main": [
[
{
"node": "e1fd4754-75cb-4d44-bc7b-01d1eddfdb75",
"type": "main",
"index": 0
},
{
"node": "c95360a1-81f7-4bf9-81c9-bcac1d918ad1",
"type": "main",
"index": 0
}
]
]
},
"cd787fbd-d7f6-411a-bd98-af776eeb94b3": {
"main": [
[
{
"node": "d3c62700-d7d3-4152-a1eb-8fd4d5b7278d",
"type": "main",
"index": 0
}
]
]
},
"9ce6eceb-d170-4ee4-897f-426f9adbf48b": {
"ai_tool": [
[
{
"node": "944e691a-6333-4382-8f96-2e85dd199c54",
"type": "ai_tool",
"index": 0
}
]
]
},
"f70fac9a-d478-42a0-a14b-5ff6cf864c57": {
"main": [
[
{
"node": "c95360a1-81f7-4bf9-81c9-bcac1d918ad1",
"type": "main",
"index": 1
}
]
]
},
"dcb79c8f-07ab-4690-ac10-c2aa7d0f08d9": {
"main": [
[
{
"node": "c95360a1-81f7-4bf9-81c9-bcac1d918ad1",
"type": "main",
"index": 1
}
],
[
{
"node": "f70fac9a-d478-42a0-a14b-5ff6cf864c57",
"type": "main",
"index": 0
}
]
]
},
"0353855d-d641-4bec-97cc-c8e53be5d9b6": {
"ai_outputParser": [
[
{
"node": "944e691a-6333-4382-8f96-2e85dd199c54",
"type": "ai_outputParser",
"index": 0
}
]
]
},
"d3c62700-d7d3-4152-a1eb-8fd4d5b7278d": {
"main": [
[
{
"node": "a28aeee2-3422-4455-a39a-84ebc0d57efd",
"type": "main",
"index": 0
}
]
]
},
"5a6e8cd3-c68d-4f24-9f3d-d55a8a0a9267": {
"main": [
[
{
"node": "013f6548-399d-43ad-a738-f280c0f3621a",
"type": "main",
"index": 0
}
],
[]
]
},
"0eb308dc-5756-488e-a941-c2c51b87e08b": {
"ai_tool": [
[
{
"node": "944e691a-6333-4382-8f96-2e85dd199c54",
"type": "ai_tool",
"index": 0
}
]
]
},
"e1fd4754-75cb-4d44-bc7b-01d1eddfdb75": {
"main": [
[
{
"node": "dcb79c8f-07ab-4690-ac10-c2aa7d0f08d9",
"type": "main",
"index": 0
}
]
]
},
"85c800e2-a0da-42e9-aaf7-b2665aeb0969": {
"main": [
[
{
"node": "5a6e8cd3-c68d-4f24-9f3d-d55a8a0a9267",
"type": "main",
"index": 0
}
]
]
},
"26684a88-c3d6-407f-af17-b54a25ca5da8": {
"ai_embedding": [
[
{
"node": "f70fac9a-d478-42a0-a14b-5ff6cf864c57",
"type": "ai_embedding",
"index": 0
}
]
]
},
"944e691a-6333-4382-8f96-2e85dd199c54": {
"main": [
[
{
"node": "45afcd19-cd43-47b0-b78b-8ecfda956359",
"type": "main",
"index": 0
}
]
]
},
"45afcd19-cd43-47b0-b78b-8ecfda956359": {
"main": [
[
{
"node": "85c800e2-a0da-42e9-aaf7-b2665aeb0969",
"type": "main",
"index": 0
},
{
"node": "013f6548-399d-43ad-a738-f280c0f3621a",
"type": "main",
"index": 1
}
],
[]
]
},
"b9161937-611c-46d5-b372-1a6f0186b637": {
"ai_outputParser": [
[
{
"node": "85c800e2-a0da-42e9-aaf7-b2665aeb0969",
"type": "ai_outputParser",
"index": 0
}
]
]
},
"12f3159a-6df5-4603-a9b0-5c742319e6da": {
"ai_languageModel": [
[
{
"node": "944e691a-6333-4382-8f96-2e85dd199c54",
"type": "ai_languageModel",
"index": 0
}
]
]
},
"e5ece6f2-9476-459d-a153-9c27aa0d6a9b": {
"ai_languageModel": [
[
{
"node": "85c800e2-a0da-42e9-aaf7-b2665aeb0969",
"type": "ai_languageModel",
"index": 0
}
]
]
},
"0218c8e3-21ec-4584-b51c-ac47b2f85364": {
"ai_document": [
[
{
"node": "f70fac9a-d478-42a0-a14b-5ff6cf864c57",
"type": "ai_document",
"index": 0
}
]
]
},
"43e0fc70-b7dc-4148-b3af-02af36ef4775": {
"ai_tool": [
[
{
"node": "944e691a-6333-4382-8f96-2e85dd199c54",
"type": "ai_tool",
"index": 0
}
]
]
},
"31f278bf-12b5-41ee-b200-a4f8e56796c9": {
"ai_textSplitter": [
[
{
"node": "0218c8e3-21ec-4584-b51c-ac47b2f85364",
"type": "ai_textSplitter",
"index": 0
}
]
]
},
"db69e156-1c18-48a3-9178-9f8341a73bfc": {
"ai_embedding": [
[
{
"node": "d3c62700-d7d3-4152-a1eb-8fd4d5b7278d",
"type": "ai_embedding",
"index": 0
}
]
]
},
"a28aeee2-3422-4455-a39a-84ebc0d57efd": {
"main": [
[],
[
{
"node": "944e691a-6333-4382-8f96-2e85dd199c54",
"type": "main",
"index": 0
}
]
]
},
"f1d39f85-4e48-43c6-ab33-9c30517a5c3e": {
"ai_embedding": [
[
{
"node": "0eb308dc-5756-488e-a941-c2c51b87e08b",
"type": "ai_embedding",
"index": 0
}
]
]
},
"a084c458-9089-41eb-8017-5523891e159f": {
"ai_embedding": [
[
{
"node": "e1fd4754-75cb-4d44-bc7b-01d1eddfdb75",
"type": "ai_embedding",
"index": 0
}
]
]
},
"bd998532-090f-4ec3-b3a8-c942dbd4d77c": {
"main": [
[
{
"node": "cd787fbd-d7f6-411a-bd98-af776eeb94b3",
"type": "main",
"index": 0
}
]
]
},
"8b99d8c3-eb24-436a-8f64-2eca763ca38c": {
"main": [
[
{
"node": "cd787fbd-d7f6-411a-bd98-af776eeb94b3",
"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 - 文書抽出, AI RAG検索拡張
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
GPT-4、Claude、Apify を使用したコンテンツ分析や洞察を目のとしたウェブ調査の自動化
GPT-4、Claude、Apify を使ったウェブ調査の自動化でコンテンツ分析とインサイトを行う
If
Set
Code
+
If
Set
Code
42 ノードPeter Zendzian
市場調査
n8nノードの探索（可視化リファレンスライブラリ内）
n8nノードを可視化リファレンスライブラリで探索
If
Ftp
Set
+
If
Ftp
Set
113 ノードI versus AI
その他
コンテキスト・ハイブリッドRAG AIコピー
RAGアプリケーション向けのGoogle DriveからSupabaseコンテキストベクトルデータベースへの同期
If
Set
Code
+
If
Set
Code
76 ノードMichael Taleb
AI RAG検索拡張
ペットショップ 4
ペットショップ予約AIエージェント
If
Set
Code
+
If
Set
Code
187 ノードBruno Dias
人工知能
デリバリー ハンバーガーショップ MVP
🤖 レストランと配送の自動化を支援するAI駆動型WhatsAppアシスタント
If
Set
Code
+
If
Set
Code
152 ノードBruno Dias
ドキュメントRAGとチャットアジェント：Google DriveからQdrantへ、Mistral OCR
ドキュメントRAGチャットエージェント：Google Drive→QdrantとMistral OCR
If
Set
Code
+
If
Set
Code
40 ノードDIGITAL BIZ TECH
内部Wiki
