Mengakses insight data di Katalog Universal Dataplex
Tetap teratur dengan koleksi
Simpan dan kategorikan konten berdasarkan preferensi Anda.
Insight data menawarkan cara otomatis untuk menjelajahi dan memahami data Anda. Dengan insight data, Gemini menggunakan metadata untuk membuat pertanyaan bahasa alami tentang tabel Anda dan kueri untuk menjawabnya. Hal ini membantu Anda mengungkap pola, menilai kualitas data, dan melakukan analisis statistik.
Dokumen ini menjelaskan fitur utama insight data dan cara melihatnya untuk eksplorasi data yang penuh insight.
Sebelum memulai
Insight data dibuat menggunakan
Gemini di BigQuery
dan hanya dapat dibuat di BigQuery Studio.
Pertama, siapkan Gemini di BigQuery, lalu buat insight di BigQuery.
Gemini memproses metadata Anda untuk mendapatkan insight secara eksklusif di region us-central1. Untuk mengetahui informasi selengkapnya, lihat
Lokasi penayangan Gemini.
Setelah membuat insight, Anda dapat melihatnya di Dataplex Universal Catalog.
Peran yang diperlukan
Untuk mendapatkan akses hanya baca ke insight yang dihasilkan, minta administrator Anda untuk memberi Anda peran IAM berikut:
Dataplex DataScan DataViewer (roles/dataplex.dataScanDataViewer)
di project yang berisi tabel BigQuery yang ingin Anda gunakan untuk
melihat insight.
Anda mungkin juga bisa mendapatkan izin yang diperlukan melalui
peran khusus atau
peran bawaan lainnya. Untuk melihat izin persis yang diperlukan untuk membuat insight, luaskan bagian Izin yang diperlukan:
Izin yang diperlukan
dataplex.datascans.get
dataplex.datascans.getData
Mengaktifkan API
Untuk menggunakan insight data, aktifkan API berikut di project Anda:
Saat menjelajahi tabel baru yang tidak dikenal, analis data sering kali menghadapi masalah cold start. Masalah ini sering kali melibatkan ketidakpastian tentang
struktur data, pola utama, dan insight yang relevan dalam
data, sehingga sulit untuk mulai menulis kueri.
Insight data mengatasi masalah mulai dingin dengan otomatis membuat kueri natural language dan padanan SQL-nya berdasarkan metadata tabel.
Daripada memulai
dengan editor kueri kosong, Anda dapat dengan cepat memulai eksplorasi data dengan
kueri yang bermakna dan menawarkan insight berharga. Untuk menyelidiki lebih lanjut, Anda
dapat mengajukan pertanyaan lanjutan di kanvas data.
Contoh menjalankan insight
Pertimbangkan tabel bernama telco_churn dengan metadata berikut:
Nama kolom
Jenis
CustomerID
STRING
Gender
STRING
Masa bakti
INT64
InternetService
STRING
StreamingTV
STRING
OnlineBackup
STRING
Kontrak
STRING
TechSupport
STRING
PaymentMethod
STRING
MonthlyCharges
FLOAT
Churn
BOOLEAN
Berikut adalah beberapa contoh kueri yang dihasilkan insight data
untuk tabel ini:
Identifikasi pelanggan yang telah berlangganan semua layanan premium dan telah menjadi pelanggan selama lebih dari 50 bulan.
Gemini beroperasi secara global, jadi Anda tidak dapat membatasi pemrosesan data ke wilayah tertentu. Untuk mempelajari lebih lanjut lokasi tempat Gemini in BigQuery memproses data, lihat Lokasi penayangan Gemini.
Batasan
Insight data tersedia untuk tabel BigQuery, tabel BigLake, tabel eksternal, dan tampilan.
Untuk pelanggan multicloud, data dari cloud lain tidak tersedia.
Insight data tidak mendukung jenis kolom Geo atau JSON.
Eksekusi insight tidak menjamin presentasi kueri setiap saat. Untuk meningkatkan kemungkinan menghasilkan kueri yang lebih menarik, buat ulang insight di BigQuery Studio.
[[["Mudah dipahami","easyToUnderstand","thumb-up"],["Memecahkan masalah saya","solvedMyProblem","thumb-up"],["Lainnya","otherUp","thumb-up"]],[["Sulit dipahami","hardToUnderstand","thumb-down"],["Informasi atau kode contoh salah","incorrectInformationOrSampleCode","thumb-down"],["Informasi/contoh yang saya butuhkan tidak ada","missingTheInformationSamplesINeed","thumb-down"],["Masalah terjemahan","translationIssue","thumb-down"],["Lainnya","otherDown","thumb-down"]],["Terakhir diperbarui pada 2025-08-21 UTC."],[[["\u003cp\u003eData insights leverage Gemini to automatically generate natural language questions and SQL queries based on a table's metadata, aiding in data exploration and analysis.\u003c/p\u003e\n"],["\u003cp\u003eTo use data insights, you must first set up Gemini in BigQuery, enable the required APIs (Dataplex, BigQuery, and Gemini for Google Cloud), and have the appropriate IAM roles, like \u003ccode\u003eDataplex DataScan DataViewer\u003c/code\u003e, for read-only access.\u003c/p\u003e\n"],["\u003cp\u003eData insights addresses the cold-start problem for data analysts by providing meaningful queries that uncover patterns, assess data quality, and offer valuable insights, which can be further explored in data canvas.\u003c/p\u003e\n"],["\u003cp\u003eInsights can be viewed in Dataplex by searching for the BigQuery table and selecting the "Insights" tab, but if they are not present, they need to be generated in BigQuery Studio.\u003c/p\u003e\n"],["\u003cp\u003eData insights has certain limitations, such as not supporting \u003ccode\u003eGeo\u003c/code\u003e or \u003ccode\u003eJSON\u003c/code\u003e column types and being available only for BigQuery tables, BigLake tables, external tables, and views.\u003c/p\u003e\n"]]],[],null,["Data insights offers an automated way to explore and understand\nyour data. With data insights, Gemini in BigQuery uses\nmetadata to generate natural language questions about your table and the queries\nto answer them. This helps you uncover patterns, assess data\nquality, and perform statistical analysis.\n\nThis document describes the key features of data insights and how to view them\nfor insightful data exploration.\n| **Note:** To give feedback for this feature, send email to [dataplex-data-insights-help@google.com](mailto:dataplex-data-insights-help@google.com).\n\nBefore you begin\n\nData insights are generated using\n[Gemini in BigQuery](/gemini/docs/bigquery/overview)\nand can only be generated in BigQuery Studio.\nFirst\n[set up Gemini in BigQuery](/gemini/docs/bigquery/set-up-gemini),\nthen\n[generate insights in BigQuery](/bigquery/docs/data-insights#insights-bigquery-table).\nGemini processes your metadata for insights exclusively in the\n`us-central1` region. For more information, see\n[Gemini serving locations](/gemini/docs/locations).\nAfter you generate insights, you can view them in Dataplex Universal Catalog.\n| **Note** : Gemini in BigQuery is part of Gemini for Google Cloud and doesn't support the same compliance and security offerings as BigQuery. You should only set up Gemini in BigQuery for BigQuery projects that don't require [compliance offerings that aren't supported by Gemini for Google Cloud](/gemini/docs/discover/certifications). For information about how to turn off or prevent access to Gemini in BigQuery, see [Turn off Gemini for Google Cloud products](/gemini/docs/turn-off-gemini).\n\nRequired roles\n\nTo get read-only access to the generated insights, ask your administrator\nto grant you the following IAM role:\n\n- [Dataplex DataScan DataViewer](/iam/docs/understanding-roles#dataplex.dataScanDataViewer) (`roles/dataplex.dataScanDataViewer`) on the project containing the BigQuery tables for which you want to view insights.\n\nFor more information about granting roles, see\n[Manage access to projects, folders, and organizations](/iam/docs/granting-changing-revoking-access).\n\nYou might also be able to get the required permissions through\n[custom roles](/iam/docs/creating-custom-roles) or other\n[predefined roles](/iam/docs/understanding-roles). To see the exact permissions\nthat are required generate insights, expand the **Required permissions**\nsection: \n\nRequired permissions\n\n- `dataplex.datascans.get`\n- `dataplex.datascans.getData`\n\nEnable APIs **Important:** To enable any API in your project, ask your administrator to grant you the `serviceusage.services.enable` permission on your project.\n\nTo use data insights, enable the following APIs in your project:\nDataplex API, BigQuery API, and Gemini for Google Cloud API.\n\n\n[Enable the APIs](https://console.cloud.google.com/flows/enableapi?apiid=dataplex.googleapis.com,bigquery.googleapis.com,cloudaicompanion.googleapis.com)\n\nFor more information about enabling the Gemini for Google Cloud API, see\n[Enable the Gemini for Google Cloud API in a Google Cloud project](/gemini/docs/discover/set-up-gemini#enable-api).\n\nAbout data insights\n\nWhen exploring a new, unfamiliar table, data analysts often face the cold-start\nproblem. The problem often involves uncertainties about\nthe data structure, key patterns, and relevant insights in\nthe data, making it hard to get started writing queries.\n\nData insights\naddresses the cold-start problem by automatically generating natural language\nqueries and their SQL equivalents based on a table's metadata.\nRather than starting\nwith an empty query editor, you can quickly start data exploration with\nmeaningful queries that offer valuable insights. To investigate further, you\ncan ask follow-up questions in [data canvas](/bigquery/docs/data-canvas).\n\nExample of an insights run\n\nConsider a table called `telco_churn` with the following metadata:\n\n| Field name | Type |\n|-----------------|-----------|\n| CustomerID | `STRING` |\n| Gender | `STRING` |\n| Tenure | `INT64` |\n| InternetService | `STRING` |\n| StreamingTV | `STRING` |\n| OnlineBackup | `STRING` |\n| Contract | `STRING` |\n| TechSupport | `STRING` |\n| PaymentMethod | `STRING` |\n| MonthlyCharges | `FLOAT` |\n| Churn | `BOOLEAN` |\n\nThe following are some of the sample queries that data insights\ngenerates for this table:\n\n- Identify customers who have subscribed to all premium services and have been\n a customer for more than 50 months.\n\n SELECT\n CustomerID,\n Contract,\n Tenure\n FROM\n agentville_datasets.telco_churn\n WHERE\n OnlineBackup = 'Yes'\n AND TechSupport = 'Yes'\n AND StreamingTV = 'Yes'\n AND Tenure \u003e 50;\n\n- Identify which internet service has the most churned customers.\n\n SELECT\n InternetService,\n COUNT(DISTINCT CustomerID) AS total_customers\n FROM\n agentville_datasets.telco_churn\n WHERE\n Churn = TRUE\n GROUP BY\n InternetService\n ORDER BY\n total_customers DESC\n LIMIT 1;\n\n- Identify churn rates by segment among high-value customers.\n\n SELECT\n Contract,\n InternetService,\n Gender,\n PaymentMethod,\n COUNT(DISTINCT CustomerID) AS total_customers,\n SUM(CASE WHEN Churn = TRUE THEN 1 ELSE 0 END) AS churned_customers,\n (SUM(CASE WHEN Churn = TRUE THEN 1 ELSE 0 END) / COUNT(DISTINCT CustomerID))\n * 100 AS churn_rate\n FROM\n agentville_datasets.telco_churn\n WHERE\n MonthlyCharges \u003e 100\n GROUP BY\n Contract,\n InternetService,\n Gender,\n PaymentMethod;\n\nView insights\n\nTo view insights for a BigQuery table, access the\ntable entry in Dataplex Universal Catalog using Dataplex Universal Catalog Search.\n\n1. In the Google Cloud console, go to the Dataplex Universal Catalog **Search** page.\n\n [Go to Search](https://console.cloud.google.com/dataplex)\n2. Search for the table entry in Dataplex Universal Catalog.\n\n3. Click the **Insights** tab. If the tab is empty, it means that the insights\n for this table are not generated yet. You can\n [generate data insights](/bigquery/docs/data-insights) in BigQuery Studio.\n\nPricing\n\nFor details about pricing for this feature, see\n[Gemini in BigQuery pricing overview](/products/gemini/pricing#gemini-in-bigquery-pricing).\n\nQuotas and limits\n\nFor information about quotas and limits for this feature, see\n[Quotas for Gemini in BigQuery](/gemini/docs/quotas#bigquery).\n\nLocations\n\nGemini operates globally, so you can't\nrestrict data processing to a specific region. To learn more about locations where\nGemini in BigQuery\nprocesses data, see [Gemini serving locations](/gemini/docs/locations).\n\nLimitations\n\n- Data insights is available for BigQuery tables, BigLake tables, external tables, and views.\n- For multicloud customers, data from other clouds isn't available.\n- Data insights doesn't support `Geo` or `JSON` column types.\n- Insights runs don't guarantee the presentation of queries every time. To increase the likelihood of generating more engaging queries, regenerate insights in BigQuery Studio.\n\nWhat's next\n\n- Learn how to [generate insights in BigQuery](/bigquery/docs/data-insights#insights-bigquery-table).\n- Learn how to [generate a data profiling scan](/dataplex/docs/use-data-profiling).\n- Learn how to [write queries with Gemini assistance in BigQuery](/bigquery/docs/write-sql-gemini)."]]
