Description

The application takes a list of categories in JSON format, passes it to the Gemini API (AI API), and outputs a JSON containing the three most popular attributes for every subcategory in the category structure.

Prerequisites

• .NET Core SDK 8

Configuration

Before running the application, you need to configure the API key. Follow these steps:

1. Create the apikey from this link https://aistudio.google.com/app/apikey. You can select any relevant project to create the API key.

2. Now in the project folder, locate the appsettings.json file within the CategoryProcessor directory.

3. Open appsettings.json and find the "GeminiAi" section.

4. Add your API key by replacing the empty string in the ApiKey field. The configuration should look like this:

 {
   "GeminiAi": {
     "ApiKey": "your-api-key-here",
     "Url": "https://generativelanguage.googleapis.com/v1beta/models/gemini-pro:generateContent"
   }
 }

Replace "your-api-key-here" with your actual API key.

Save the file after making the changes.

Execution

To restore dependencies and run the application, use the following commands in the command window:

dotnet restore
dotnet run --launch-profile https

Usage

1. Once the application runs successfully, in the web browser, go to the URL: https://localhost:7046/.
2. Enter the list of categories in the input field. The input json to test is present in input_data.txt file
3. Click on the submit button.
4. The list of attributes will be populated after a few seconds in the output field.

Note

An empty list of attributes may be returned for some categories because the Gemini API can become exhausted due to too many requests in a short time.

Troubleshoot

For troubleshooting, you can view the logs in the command window where you ran the application.