Skillset Tutorial in REST

Browse code

This sample uses the Azure AI Search REST APIs to create an indexer, data source, index, and skillset that applies AI processing to Azure blobs during indexing.

Skillset steps include OCR, language detection, entity recognition, and key phrase extraction. The sample demo files consist of multiple content types ranging from image-only, text-only, and rich application files with embedded images and text. The end result is searchable index that you can query for new and enriched content.

This collection is featured in the Tutorial: Use REST and AI to generate searchable content from Azure blobs. If you have trouble with the steps in this readme, check the tutorial for more detailed instructions.

Prerequisites

• Visual Studio Code with a REST client.

• Azure AI Search. Create or find an existing Azure AI Search resource under your current subscription.

• Azure Storage account

• Sample data files (mixed media)

Set up the data

1. Download the sample data files.

2. In Azure portal, in Azure Storage, create a Blob container and upload the sample data files.

3. Make a note of the blob container name.

Get connection information

Gather connection information used on the requests. You can find this information in the Azure portal. Save it in Notepad or another temporary location.

1. In Azure Storage, select Access keys on the left. Copy one of the connection strings. It should be in this format: DefaultEndpointsProtocol=https;AccountName=<YOUR-STORAGE-ACCOUNT>;AccountKey=<YOUR-ACCESS-KEY>;

2. In Azure AI Search, get the endpoint (for example, https://demo-svc.search.windows.net). Next, select Keys on the left and copy one of admin keys.

Set up variables

1. Clone or download this sample repository.

2. Open skillset-tutorial.rest in Visual Studio Code.

3. Paste in the variables you collected earlier:

   • In @baseUrl, enter the search endpoint.
   • In @apiKey, enter the admin API key of your search service.
   • In @storageConnectionString, enter the full access connection string for your Azure Storage account.
   • In @blobContainer, enter the name of the blob container that stores the clinical trials documents.

Create objects and queryu the index

1. Send each request to create a data source, indexer, skillset, and index used in this example.

2. The last set of requests query the search index. For more information about this sample, see Tutorial: Use skillsets to generate searchable content in Azure AI Search.

Next steps

You can learn more about Azure AI Search on the official documentation site.