Siapkan project Google Cloud , aktifkan Aplikasi AI, dan siapkan kontrol akses untuk project Anda. Anda dapat menggunakan project Google Cloud yang ada jika sudah menggunakannya.
Siapkan data Anda untuk diimpor ke Vertex AI Search.
Aplikasi rekomendasi media memerlukan dua jenis data:
Data media terstruktur. Upload informasi metadata tentang konten
media Anda, seperti judul, deskripsi, dan URI ke lokasi media Anda.
Vertex AI Search menyediakan skema standar untuk media.
Atau, Anda dapat menggunakan skema Anda sendiri.
Peristiwa pengguna. Merekam peristiwa pengguna diperlukan untuk rekomendasi media.
Peristiwa pengguna diperlukan untuk melatih aplikasi Anda dan membuat rekomendasi.
Buat aplikasi dan penyimpanan data, lalu impor data media dan peristiwa pengguna.
Cara mengimpor data media bergantung pada tempat Anda mengimpornya. Misalnya, jika data Anda berada di Cloud Storage, Anda dapat mengimpornya menggunakan
konsol atau API dengan memberikan lokasi bucket data Anda.
Anda dapat memperbarui setelan kolom agar dapat difilter dan memfilter
hasil rekomendasi menggunakan kolom tersebut.
Tindakan
Agar Vertex AI Search dapat menggunakan kolom tertentu untuk memfilter rekomendasi, tetapkan kolom tersebut sebagai dapat difilter. Lihat Mengonfigurasi setelan kolom.
Anda dapat melihat pratinjau rekomendasi untuk memeriksa apakah rekomendasi Anda
muncul seperti yang diharapkan.
Tindakan
Untuk melihat pratinjau rekomendasi, gunakan konsol Aplikasi AI atau
API.
Konsol. Gunakan halaman Pratinjau konsol untuk melihat pratinjau rekomendasi Anda. Lihat petunjuk Konsol untuk mengetahui jenis data yang
digunakan aplikasi Anda di Mendapatkan
rekomendasi.
API. Jika Anda mengintegrasikan panggilan API ke dalam aplikasi, lakukan panggilan
API untuk melihat pratinjau rekomendasi Anda. Lihat petunjuk REST
untuk jenis data yang digunakan aplikasi Anda di Mendapatkan
rekomendasi.
Jika Anda puas dengan versi pratinjau aplikasi rekomendasi media,
bagikan kepada pengguna dengan men-deploy-nya ke situs Anda.
Tindakan
Untuk men-deploy aplikasi rekomendasi, integrasikan panggilan API ke server atau
aplikasi Anda. Untuk informasi selengkapnya tentang cara melakukan panggilan API, lihat petunjuk REST
untuk jenis data yang digunakan aplikasi Anda di Mendapatkan
rekomendasi.
Anda dapat mengelola aplikasi untuk memastikan data terbaru dan yang diperlukan tersedia
di penyimpanan data Anda.
Aplikasi rekomendasi media otomatis disesuaikan setiap 3 bulan. Namun, jika
penyimpanan data Anda mengalami perubahan yang signifikan atau jika Anda melakukan impor massal peristiwa
pengguna, Anda harus melatih ulang aplikasi secara manual.
[[["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 2025-08-19 UTC."],[[["\u003cp\u003eThis page outlines the steps to implement a media recommendations app using Vertex AI Search, from initial setup to deployment and maintenance.\u003c/p\u003e\n"],["\u003cp\u003ePreparing your data involves providing structured media data and recording user events, which are both crucial for app training and generating recommendations.\u003c/p\u003e\n"],["\u003cp\u003eCreating the app requires making a data store, importing your media content, setting up historical data, and setting up live tracking for user interactions.\u003c/p\u003e\n"],["\u003cp\u003eCustomization of recommendations is achieved by configuring field settings for filtering, boosting, burying, diversifying, and demoting content.\u003c/p\u003e\n"],["\u003cp\u003eOnce implemented, you can preview, deploy, and maintain the recommendations app, including retraining it after significant data changes or bulk imports of user events.\u003c/p\u003e\n"]]],[],null,["# Media recommendations checklist\n\nThis page provides a checklist of the steps required to implement a media recommendations app.\n\n\u003cbr /\u003e\n\nIf you're new to Vertex AI Search, consider following the [Get started\nwith media recommendations](/generative-ai-app-builder/docs/try-media-recommendations) quickstart\ntutorial to create a sample app.\n\n\n### [Set up a Google Cloud\nproject](#set-up-project)\n\nSet up a Google Cloud project, turn on AI Applications, and set up access\ncontrol for your project. You can use an existing Google Cloud project if you\nhave one already.\n\n### Actions\n\n1. Complete the steps in [Before you begin](/generative-ai-app-builder/docs/before-you-begin). \n\n### [Prepare your data](#prepare-data)\n\nPrepare your data for importing to Vertex AI Search.\n\nMedia recommendations apps require two types of data:\n\n- **Structured media data.** Upload metadata information about your media\n content, such as titles, descriptions, and URIs to the location of your media.\n Vertex AI Search provides a predefined schema for media.\n Alternatively, you can use your own schema.\n\n- **User events.** Recording user events is required for media recommendations.\n User events are required to train your app and to generate recommendations.\n\n### Actions\n\n1. Review information about media data and data stores and prepare your data\n according to the required schemas and fields in [About media documents and\n data stores](/generative-ai-app-builder/docs/media-documents). If you are\n using your own schema, also see [Example schema as a JSON\n object](/generative-ai-app-builder/docs/provide-schema#example-schema) and\n [structured data](/generative-ai-app-builder/docs/prepare-data#structured).\n\n2. Review media user event requirements in [About user\n events](/generative-ai-app-builder/docs/media-user-events).\n\n### [Create your app and import\nyour data](#import-data)\n\nCreate an app and data store and then import your media data and user events.\n\nHow you import media data depends on where you're importing it from. For\nexample, if your data is in Cloud Storage, you can import it using the\nconsole or the API by providing the bucket location of your data.\n\n### Actions\n\n1. [Create a media data store](/generative-ai-app-builder/docs/create-data-store-media).\n\n2. [Create a media recommendations\n app](/generative-ai-app-builder/docs/create-app-media#recs).\n\n3. [Bulk import historical user events](/generative-ai-app-builder/docs/import-user-events) so that\n your app can start training.\n\n4. [Set up real-time user event recording](/generative-ai-app-builder/docs/record-user-events).\n\n### [Configure your\nrecommendations](#configure-settings)\n\nYou can update field settings to make them filterable and filter your\nrecommendations results using those fields.\n\n### Actions\n\n1. To let Vertex AI Search use a certain field for filtering\n recommendations, set that field as filterable. See [Configure field\n settings](/generative-ai-app-builder/docs/configure-field-settings).\n\n2. [Filter recommendations](/generative-ai-app-builder/docs/filter-recommendations).\n\n3. [Boost and bury recommendations](/generative-ai-app-builder/docs/media-recommendations-boost-bury).\n\n4. [Diversify media recommendations](/generative-ai-app-builder/docs/diversify-recommendations).\n\n5. Demote content if a user has recently viewed it or if the content is old. See\n [Demote media recommendations](/generative-ai-app-builder/docs/demote-recommendations).\n\n### [Preview\nrecommendations](#preview-results)\n\nYou can preview your recommendations to check if your recommendations are\nappearing as expected.\n\n### Actions\n\n1. To preview your recommendations, use the AI Applications console or\n the API.\n\n - **Console.** Use the console **Preview** page to preview your\n recommendations. See the **Console** instructions for the kind of data that\n your app uses in [Get\n recommendations](/generative-ai-app-builder/docs/preview-recommendations).\n\n - **API** . If you're integrating API calls into your application, make API\n calls to preview your recommendations. See the **REST**\n instructions for the kind of data that your app uses in [Get\n recommendations](/generative-ai-app-builder/docs/preview-recommendations).\n\n### [Deploy your\nrecommendations app](#preview-results)\n\nWhen you're happy with the preview version of your media recommendations app,\nshare it with your users by deploying it to your website.\n\n### Actions\n\n1. To deploy your recommendations app, integrate API calls into your server or\n applications. For more information about making API calls, see the **REST**\n instructions for the kind of data that your app uses in [Get\n recommendations](/generative-ai-app-builder/docs/preview-recommendations).\n\n For client library resources, see [AI Applications client\nlibraries](/generative-ai-app-builder/docs/libraries). \n\n### [Maintain your app](#maintain-data)\n\nYou can maintain your app to ensure that latest and necessary data is available\nin your data store.\n\nMedia recommendations apps are automatically tuned every 3 months. However, if\nyour data store undergoes significant change or if you do a bulk import of user\nevents, you should manually retrain your app.\n\n### Actions\n\n1. [Check the quality of your media data](/generative-ai-app-builder/docs/check-media-data-quality).\n\n2. [Refresh your structured data](/generative-ai-app-builder/docs/refresh-data#refresh-structured).\n\n3. [Train and tune media apps](/generative-ai-app-builder/docs/train-tune-media)."]]
