IELTS Assessment Using Speechsuper API

Step 1: Obtaining Credentials

To use the Speechsuper API, you need an appKey and secretKey. Follow these steps to obtain them:

1. Sign Up:

   • Go to the Speechsuper website.
   • Register for an account if you don't already have one.

2. Create an Application:

   • After logging in, navigate to the API management section or dashboard.
   • Create a new application if one does not already exist. This will allow you to generate the necessary API keys.

3. Obtain appKey and secretKey:

   • Once your application is created, you should see your appKey and secretKey in the application details.
   • Keep these credentials secure as they are used to authenticate your API requests.

Step 2: Setting Up Your Configuration

Create a config.py file in your project directory to store your credentials. This file should look like this:

# config.py
appKey = "your_app_key_here"
secretKey = "your_secret_key_here"

Replace "your_app_key_here" and "your_secret_key_here" with the values you obtained from Speechsuper.

Step 3: Running the SpeechSuper Code

To see how the API can be used, follow these steps:

1. Read Through the SpeechSuper.ipynb Notebook:

   • The notebook provides a detailed example of how to use the Speechsuper API for IELTS assessment, including audio format conversion and result extraction.

2. Modify the Example Code:

   • In the notebook, locate the section where the API is called. Change the following variables to match your test case:

   q = "What are your hobbies?"  # Replace with your actual question
   audio = "converted_What are your hobbies.wav"  # Replace with your actual audio file path
   result = get_IELTS_assessment_report(input_audio=audio, question_ielts=q)

   Ensure that the audio variable points to a valid audio file path and the q variable contains the appropriate question text.

Step 4: Speech Performance Analysis

After obtaining the result.json file from Step 3, follow these steps to analyze the speech performance:

1. Process the Results:

   Use the run_notebook.py script to process the result.json file. This script will:

   • Read the assessment results from result.json.
   • Extract performance data for each IELTS speaking category: Fluency, Grammar, Lexical Resource, and Pronunciation.
   • Convert the performance data into natural language summaries and save them to appropriate files.

2. Generated Outputs:

   The script will produce the following files:

   • Fluency_Input_data_to_prompt.txt: Contains natural language summaries of fluency performance.
   • Lexical_Input_data_to_prompt.txt: Contains natural language summaries of lexical resource performance.
   • Grammar_Input_data_to_prompt.txt: Contains natural language summaries of grammar performance.
   • Pronunciation_Input_data_to_prompt.json: Contains detailed JSON analysis of pronunciation performance.

Step 5: Generating Tailor-Made Feedback and Recommendations

1. Prepare System Prompts:

   • In OpenAI assistance, set the following system prompts:
     • System_prompt_Fluency.txt
     • System_prompt_Grammar.txt
     • System_prompt_Lexical.txt
     • System_prompt_Pronunciation.txt

2. Match Performance Data:

   • Load the performance data files into the corresponding OpenAI assistance:

   • Generate tailored feedback and recommendations based on the performance data and system prompts.

Acknowledgments

• Speechsuper: For providing the API used for IELTS assessment.
• OpenAI GPT: For enabling the generation of detailed feedback and recommendations.
• Pydub: For audio format conversion and processing.
• Requests: For handling API interactions and data retrieval.