Example: Create a custom skill using the Bing Entity Search API

In this example, learn how to create a web API custom skill. This skill will accept locations, public figures, and organizations, and return descriptions for them. The example uses an Azure Function to wrap the Bing Entity Search API so that it implements the custom skill interface.

Prerequisites

  • Read about custom skill interface article if you aren't familiar with the input/output interface that a custom skill should implement.

  • Create a Bing Search resource through the Azure portal. A free tier is available and sufficient for this example.

  • Install Visual Studio or later.

Create an Azure Function

Although this example uses an Azure Function to host a web API, it isn't required. As long as you meet the interface requirements for a cognitive skill, the approach you take is immaterial. Azure Functions, however, make it easy to create a custom skill.

Create a project

  1. In Visual Studio, select New > Project from the File menu.

  2. Choose Azure Functions as the template and select Next. Type a name for your project, and select Create. The function app name must be valid as a C# namespace, so don't use underscores, hyphens, or any other special characters.

  3. Select a framework that has long term support.

  4. Choose HTTP Trigger for the type of function to add to the project.

  5. Choose Function for the authorization level.

  6. Select Create to create the function project and HTTP triggered function.

Add code to call the Bing Entity API

Visual Studio creates a project with boilerplate code for the chosen function type. The FunctionName attribute on the method sets the name of the function. The HttpTrigger attribute specifies that the function is triggered by an HTTP request.

Replace the contents of Function1.cs with the following code:

using System;
using System.Collections.Generic;
using System.IO;
using System.Linq;
using System.Net.Http;
using System.Threading.Tasks;
using Microsoft.AspNetCore.Mvc;
using Microsoft.Azure.WebJobs;
using Microsoft.Azure.WebJobs.Extensions.Http;
using Microsoft.AspNetCore.Http;
using Microsoft.Extensions.Logging;
using Newtonsoft.Json;

namespace SampleSkills
{
    /// <summary>
    /// Sample custom skill that wraps the Bing entity search API to connect it with a 
    /// AI enrichment pipeline.
    /// </summary>
    public static class BingEntitySearch
    {
        #region Credentials
        // IMPORTANT: Make sure to enter your credential and to verify the API endpoint matches yours.
        static readonly string bingApiEndpoint = "https://api.bing.microsoft.com/v7.0/entities";
        static readonly string key = "<enter your api key here>";  
        #endregion

        #region Class used to deserialize the request
        private class InputRecord
        {
            public class InputRecordData
            {
                public string Name { get; set; }
            }

            public string RecordId { get; set; }
            public InputRecordData Data { get; set; }
        }

        private class WebApiRequest
        {
            public List<InputRecord> Values { get; set; }
        }
        #endregion

        #region Classes used to serialize the response

        private class OutputRecord
        {
            public class OutputRecordData
            {
                public string Name { get; set; } = "";
                public string Description { get; set; } = "";
                public string Source { get; set; } = "";
                public string SourceUrl { get; set; } = "";
                public string LicenseAttribution { get; set; } = "";
                public string LicenseUrl { get; set; } = "";
            }

            public class OutputRecordMessage
            {
                public string Message { get; set; }
            }

            public string RecordId { get; set; }
            public OutputRecordData Data { get; set; }
            public List<OutputRecordMessage> Errors { get; set; }
            public List<OutputRecordMessage> Warnings { get; set; }
        }

        private class WebApiResponse
        {
            public List<OutputRecord> Values { get; set; }
        }
        #endregion

        #region Classes used to interact with the Bing API
        private class BingResponse
        {
            public BingEntities Entities { get; set; }
        }
        private class BingEntities
        {
            public BingEntity[] Value { get; set; }
        }

        private class BingEntity
        {
            public class EntityPresentationinfo
            {
                public string[] EntityTypeHints { get; set; }
            }

            public class License
            {
                public string Url { get; set; }
            }

            public class ContractualRule
            {
                public string _type { get; set; }
                public License License { get; set; }
                public string LicenseNotice { get; set; }
                public string Text { get; set; }
                public string Url { get; set; }
            }

            public ContractualRule[] ContractualRules { get; set; }
            public string Description { get; set; }
            public string Name { get; set; }
            public EntityPresentationinfo EntityPresentationInfo { get; set; }
        }
        #endregion

        #region The Azure Function definition

        [FunctionName("EntitySearch")]
        public static async Task<IActionResult> Run(
            [HttpTrigger(AuthorizationLevel.Function, "post", Route = null)] HttpRequest req,
            ILogger log)
        {
            log.LogInformation("Entity Search function: C# HTTP trigger function processed a request.");

            var response = new WebApiResponse
            {
                Values = new List<OutputRecord>()
            };

            string requestBody = new StreamReader(req.Body).ReadToEnd();
            var data = JsonConvert.DeserializeObject<WebApiRequest>(requestBody);

            // Do some schema validation
            if (data == null)
            {
                return new BadRequestObjectResult("The request schema does not match expected schema.");
            }
            if (data.Values == null)
            {
                return new BadRequestObjectResult("The request schema does not match expected schema. Could not find values array.");
            }

            // Calculate the response for each value.
            foreach (var record in data.Values)
            {
                if (record == null || record.RecordId == null) continue;

                OutputRecord responseRecord = new OutputRecord
                {
                    RecordId = record.RecordId
                };

                try
                {
                    responseRecord.Data = GetEntityMetadata(record.Data.Name).Result;
                }
                catch (Exception e)
                {
                    // Something bad happened, log the issue.
                    var error = new OutputRecord.OutputRecordMessage
                    {
                        Message = e.Message
                    };

                    responseRecord.Errors = new List<OutputRecord.OutputRecordMessage>
                    {
                        error
                    };
                }
                finally
                {
                    response.Values.Add(responseRecord);
                }
            }

            return (ActionResult)new OkObjectResult(response);
        }

        #endregion

        #region Methods to call the Bing API
        /// <summary>
        /// Gets metadata for a particular entity based on its name using Bing Entity Search
        /// </summary>
        /// <param name="entityName">The name of the entity to extract data for.</param>
        /// <returns>Asynchronous task that returns entity data. </returns>
        private async static Task<OutputRecord.OutputRecordData> GetEntityMetadata(string entityName)
        {
            var uri = bingApiEndpoint + "?q=" + entityName + "&mkt=en-us&count=10&offset=0&safesearch=Moderate";
            var result = new OutputRecord.OutputRecordData();

            using (var client = new HttpClient())
            using (var request = new HttpRequestMessage {
                Method = HttpMethod.Get,
                RequestUri = new Uri(uri)
            })
            {
                request.Headers.Add("Ocp-Apim-Subscription-Key", key);

                HttpResponseMessage response = await client.SendAsync(request);
                string responseBody = await response?.Content?.ReadAsStringAsync();

                BingResponse bingResult = JsonConvert.DeserializeObject<BingResponse>(responseBody);
                if (bingResult != null)
                {
                    // In addition to the list of entities that could match the name, for simplicity let's return information
                    // for the top match as additional metadata at the root object.
                    return AddTopEntityMetadata(bingResult.Entities?.Value);
                }
            }

            return result;
        }

        private static OutputRecord.OutputRecordData AddTopEntityMetadata(BingEntity[] entities)
        {
            if (entities != null)
            {
                foreach (BingEntity entity in entities.Where(
                    entity => entity?.EntityPresentationInfo?.EntityTypeHints != null
                        && (entity.EntityPresentationInfo.EntityTypeHints[0] == "Person"
                            || entity.EntityPresentationInfo.EntityTypeHints[0] == "Organization"
                            || entity.EntityPresentationInfo.EntityTypeHints[0] == "Location")
                        && !String.IsNullOrEmpty(entity.Description)))
                {
                    var rootObject = new OutputRecord.OutputRecordData
                    {
                        Description = entity.Description,
                        Name = entity.Name
                    };

                    if (entity.ContractualRules != null)
                    {
                        foreach (var rule in entity.ContractualRules)
                        {
                            switch (rule._type)
                            {
                                case "ContractualRules/LicenseAttribution":
                                    rootObject.LicenseAttribution = rule.LicenseNotice;
                                    rootObject.LicenseUrl = rule.License.Url;
                                    break;
                                case "ContractualRules/LinkAttribution":
                                    rootObject.Source = rule.Text;
                                    rootObject.SourceUrl = rule.Url;
                                    break;
                            }
                        }
                    }

                    return rootObject;
                }
            }

            return new OutputRecord.OutputRecordData();
        }
        #endregion
    }
}

Make sure to enter your own key value in the key constant based on the key you got when signing up for the Bing entity search API.

Test the function from Visual Studio

Press F5 to run the program and test function behaviors. In this case, we'll use the function below to look up two entities. Use a REST client to issue a call like the one shown below:

POST https://localhost:7071/api/EntitySearch

Request body

{
    "values": [
        {
            "recordId": "e1",
            "data":
            {
                "name":  "Pablo Picasso"
            }
        },
        {
            "recordId": "e2",
            "data":
            {
                "name":  "Microsoft"
            }
        }
    ]
}

Response

You should see a response similar to the following example:

{
    "values": [
        {
            "recordId": "e1",
            "data": {
                "name": "Pablo Picasso",
                "description": "Pablo Ruiz Picasso was a Spanish painter [...]",
                "source": "Wikipedia",
                "sourceUrl": "http://en.wikipedia.org/wiki/Pablo_Picasso",
                "licenseAttribution": "Text under CC-BY-SA license",
                "licenseUrl": "http://creativecommons.org/licenses/by-sa/3.0/"
            },
            "errors": null,
            "warnings": null
        },
        "..."
    ]
}

Publish the function to Azure

When you're satisfied with the function behavior, you can publish it.

  1. In Solution Explorer, right-click the project and select Publish. Choose Create New > Publish.

  2. If you haven't already connected Visual Studio to your Azure account, select Add an account....

  3. Follow the on-screen prompts. You're asked to specify a unique name for your app service, the Azure subscription, the resource group, the hosting plan, and the storage account you want to use. You can create a new resource group, a new hosting plan, and a storage account if you don't already have these. When finished, select Create

  4. After the deployment is complete, notice the Site URL. It is the address of your function app in Azure.

  5. In the Azure portal, navigate to the Resource Group, and look for the EntitySearch Function you published. Under the Manage section, you should see Host Keys. Select the Copy icon for the default host key.

Test the function in Azure

Now that you have the default host key, test your function as follows:

POST https://[your-entity-search-app-name].azurewebsites.net/api/EntitySearch?code=[enter default host key here]

Request Body

{
    "values": [
        {
            "recordId": "e1",
            "data":
            {
                "name":  "Pablo Picasso"
            }
        },
        {
            "recordId": "e2",
            "data":
            {
                "name":  "Microsoft"
            }
        }
    ]
}

This example should produce the same result you saw previously when running the function in the local environment.

Connect to your pipeline

Now that you have a new custom skill, you can add it to your skillset. The example below shows you how to call the skill to add descriptions to organizations in the document (this could be extended to also work on locations and people). Replace [your-entity-search-app-name] with the name of your app.

{
    "skills": [
      "[... your existing skills remain here]",  
      {
        "@odata.type": "#Microsoft.Skills.Custom.WebApiSkill",
        "description": "Our new Bing entity search custom skill",
        "uri": "https://[your-entity-search-app-name].azurewebsites.net/api/EntitySearch?code=[enter default host key here]",
          "context": "/document/merged_content/organizations/*",
          "inputs": [
            {
              "name": "name",
              "source": "/document/merged_content/organizations/*"
            }
          ],
          "outputs": [
            {
              "name": "description",
              "targetName": "description"
            }
          ]
      }
  ]
}

Here, we're counting on the built-in entity recognition skill to be present in the skillset and to have enriched the document with the list of organizations. For reference, here's an entity extraction skill configuration that would be sufficient in generating the data we need:

{
    "@odata.type": "#Microsoft.Skills.Text.V3.EntityRecognitionSkill",
    "name": "#1",
    "description": "Organization name extraction",
    "context": "/document/merged_content",
    "categories": [ "Organization" ],
    "defaultLanguageCode": "en",
    "inputs": [
        {
            "name": "text",
            "source": "/document/merged_content"
        },
        {
            "name": "languageCode",
            "source": "/document/language"
        }
    ],
    "outputs": [
        {
            "name": "organizations",
            "targetName": "organizations"
        }
    ]
},

Next steps

Congratulations! You've created your first custom skill. Now you can follow the same pattern to add your own custom functionality. Click the following links to learn more.