/essay_evaluator

Bot that evaluates ENEMs essays

Primary LanguagePython

ENEM Essay Evaluator

ENEM is Brazil's national high school exam, and its essay requires an argumentative text on a given social issue. The essay is evaluated based on structure, language, and proposed solutions.

There are 5 competencies:

  1. Formal Writing Mastery
  2. Essay Comprehension
  3. Argument Organization
  4. Argumentation Techniques
  5. Intervention Proposal

Each competency is individually scored from 0 to 200, and the scores are summed to calculate the final grade. This app is designed to analyze an ENEM essay, considering the theme, grading it, and providing detailed feedback.

Watch the video

Features

  • Essay Evaluation: Input your essay and its theme, and the app will automatically evaluate the essay across five competences.
  • LLM Integration: Utilizes OpenAI's language models to perform the evaluation.
  • Real-time Feedback: View detailed scores and explanations for each competence instantly after evaluation.
  • User-friendly Interface: Built with Streamlit, providing a simple and clean UI.

Competences Evaluated

  • Competence 1: Formal Writing Mastery (Grammar, spelling, punctuation, coherence, and vocabulary usage).
  • Competence 2: Essay Comprehension (Understanding the essay prompt and effectively using knowledge from different fields).
  • Competence 3: Argument Organization (Selecting, organizing, and interpreting information to defend a point of view).
  • Competence 4: Argumentation Mechanisms (Using linguistic tools to construct arguments).
  • Competence 5: Intervention Proposal (Proposing a solution to the problem discussed in the essay).

Project structure

  • evaluator_frontend.py: File containing the app front-end
  • graph_essay_evaluator.png: Image file with the graph of the app
  • main.py: Main file with all the logic to run the LLM
  • poetry.lock, main.cpython-312.pyc, pyproject.toml: Files from the poetry package
  • README.md: Project documentation

Installation

  1. Clone the repository:

    git clone https://github.com/nialefs/essay_evaluator.git
cd essay_evaluator

  2. Init the poetry project and install dependencies:

    poetry init
poetry shell
poetry install

  3. Set up environment variables. Create a .env file in the project root with the following contents:

    OPENAI_API_KEY=your_openai_api_key
LANGCHAIN_API_KEY=your_langchain_api_key

  4. Run your streamlit app with

    streamlit run evaluator_frontend.py

  5. Install ngrok following the steps provided in the website and add the link provided by streamlit:

    ngrok http http://192.168.1.104:8501

  6. A link will be generated and you can share with other people

Key dependencies:

  • python = "^3.12"
  • langgraph = "^0.2.35"
  • langchain-openai = "^0.2.2"
  • langchain-core = "^0.3.10"
  • python-dotenv = "^1.0.1"
  • streamlit = "^1.39.0"

Acknowledgments