BigQuery에 상호작용 로깅을 내보낼 수 있습니다.
구성되면 모든 실시간 상호작용 로깅이 BigQuery 테이블에 기록됩니다.
이를 통해 상담사를 디버그하고 개선하며 대화 데이터에서 패턴을 찾는 데 도움이 되는 고급 분석 도구를 사용할 수 있습니다.
제한사항
다음과 같은 제한사항이 적용됩니다.
대화당 최대 500회까지 내보낼 수 있습니다.
프로젝트 간 권한
Dialogflow 에이전트와 BigQuery 데이터가 동일한 프로젝트에 있지 않으면 Dialogflow Google Cloud 프로젝트와 연결된 서비스 계정에는 BigQuery Google Cloud 프로젝트의 BigQuery 데이터 세트에 대한 roles/bigquery.dataEditor IAM 권한도 있어야 합니다.
서비스 계정 형식: service-<dialogflow-project-number>@gcp-sa-dialogflow.iam.gserviceaccount.com
Dialogflow에서 내보내기를 구성하는 사용자에게 BigQuery 프로젝트에 대한 권한이 있어야 합니다. 그렇지 않으면 BigQuery 프로젝트가 Dialogflow Console에 옵션으로 표시되지 않습니다.
사용자가 Dialogflow에서 BigQuery 프로젝트를 볼 수 있으려면 BigQuery 프로젝트에 필요한 최소 권한은 resourcemanager.projects.get입니다.
또는 이 권한을 포함하지만 사용자가 BigQuery 데이터 세트에 액세스할 필요가 없는 다음 Google Cloud 사전 정의된 역할 중 하나(roles/browser 또는 roles/bigquery.metadataViewer)를 할당할 수 있습니다.
[[["이해하기 쉬움","easyToUnderstand","thumb-up"],["문제가 해결됨","solvedMyProblem","thumb-up"],["기타","otherUp","thumb-up"]],[["이해하기 어려움","hardToUnderstand","thumb-down"],["잘못된 정보 또는 샘플 코드","incorrectInformationOrSampleCode","thumb-down"],["필요한 정보/샘플이 없음","missingTheInformationSamplesINeed","thumb-down"],["번역 문제","translationIssue","thumb-down"],["기타","otherDown","thumb-down"]],["최종 업데이트: 2025-09-04(UTC)"],[[["\u003cp\u003eDialogflow interaction logs can be exported to BigQuery, enabling advanced analysis for debugging, agent improvement, and identifying conversation patterns.\u003c/p\u003e\n"],["\u003cp\u003eA maximum of 500 turns per conversation can be exported, and if the Dialogflow agent and BigQuery data are in different projects, the Dialogflow service account needs \u003ccode\u003eroles/bigquery.dataEditor\u003c/code\u003e permission on the BigQuery dataset.\u003c/p\u003e\n"],["\u003cp\u003eEach row in the exported BigQuery table represents a single conversational turn and contains columns such as project ID, agent ID, conversation name, turn number, request/response details, and additional metadata.\u003c/p\u003e\n"],["\u003cp\u003eTo configure the export, interaction logging must be enabled, a BigQuery dataset and table must be created using the provided SQL schema, and the agent settings must be updated with the dataset and table names.\u003c/p\u003e\n"],["\u003cp\u003eUsers configuring the export must have permissions on the BigQuery project, with the minimum being \u003ccode\u003eresourcemanager.projects.get\u003c/code\u003e or one of the roles \u003ccode\u003eroles/browser\u003c/code\u003e or \u003ccode\u003eroles/bigquery.metadataViewer\u003c/code\u003e.\u003c/p\u003e\n"]]],[],null,["# Interaction logging export to BigQuery\n\nYou can export\n[interaction logging](/dialogflow/cx/docs/concept/agent-settings#settings-general)\nto\n[BigQuery](/bigquery/docs).\nOnce configured, all live interaction logging is written\nto your BigQuery table.\nThis provides you with advanced analysis tools\nthat can help you debug and improve your agent\nand to discover patterns in conversation data.\n\nLimitations\n-----------\n\nThe following limitations apply:\n\n- A maximum of 500 turns can be exported for each conversation.\n\nCross project permissions\n-------------------------\n\nIf your Dialogflow agent and BigQuery data are not in the\nsame project, the [service account](/dialogflow/cx/docs/quick/setup#sa)\nassociated with your Dialogflow Google Cloud project must also have the\n`roles/bigquery.dataEditor` IAM permission for the BigQuery dataset in\nyour BigQuery Google Cloud project.\n\nService account format: `service-\u003cdialogflow-project-number\u003e@gcp-sa-dialogflow.iam.gserviceaccount.com`\n\nThe user configuring the export in Dialogflow must have permissions on\nthe BigQuery project. If they don't, the BigQuery project\nwon't appear as an option in the Dialogflow Console.\n\nThe minimum permission required on the BigQuery project in order for\nthe user to see it in Dialogflow is `resourcemanager.projects.get`.\nAlternatively, you can assign one of the following Google Cloud predefined roles\nthat include this permission but don't require the user to have access to the\nBigQuery dataset: `roles/browser` or `roles/bigquery.metadataViewer`.\n\nTable description\n-----------------\n\nEach row of the table contains one conversational turn\nwith the following columns:\n\nConfiguration\n-------------\n\nTo configure interaction logging export:\n\n1. Ensure that [interaction logging](/dialogflow/cx/docs/concept/agent-settings#settings-general) is enabled.\n2. Follow the BigQuery [dataset creation guide](/bigquery/docs/datasets#create-dataset) to create a dataset. Note the dataset name, as you will need this in the next step.\n3. Follow the BigQuery\n [table creation guide](/bigquery/docs/tables)\n to create a table with a SQL schema definition.\n Use the following SQL statement for creation:\n\n CREATE TABLE \u003cyour_dataset_name\u003e.dialogflow_bigquery_export_data(\n project_id STRING,\n agent_id STRING,\n conversation_name STRING,\n turn_position INTEGER,\n request_time TIMESTAMP,\n language_code STRING,\n request JSON,\n response JSON,\n partial_responses JSON,\n derived_data JSON,\n conversation_signals JSON,\n bot_answer_feedback JSON\n );\n\n4. Configure your\n [agent settings](/dialogflow/cx/docs/concept/agent-settings#settings-general)\n to enable BigQuery export, and to provide the dataset and table names\n created above.\n\n | **Note:** If your agent is protected by a Customer Managed Encryption Key (CMEK), you will see a warning if the destination BigQuery table is not protected by CMEK."]]
