Skip to content

paniaton/Kaggle-Automated-Essay-Checking-System

 
 

Repository files navigation

Kaggle-Automated-Essay-Checking-System

My attempt for the Kaggle AES project (https://www.kaggle.com/c/asap-aes) The linear regression model uses a Word2Vec model and custom generated heuristic features to obtain a mean-quadratic-weighted-kappa score of 0.9359.

Notebooks:

  1. CustomFeatureGeneration.ipynb - Generating custom features for the data set.
  2. Data_Exploration.ipynb - Exploring the data set and free form visualization.
  3. Linear Regression Model.ipynb - The model building and learning takes place here.

Helper functions and required library import:

./utils/helperfunctions.py
./utils/requirements.py

Libraries used for the Capstone Project:

  1. Scikit-learn 0.18.1: pip install --user --upgrade sklearn
  2. Gensim 2.1.0: pip install --user --upgrade gensim
  3. Textmining 1.0: pip install --user --upgrade textmining
  4. Grammar Check 1.3.1 : 1. pip install --upgrade 3to2 2. pip install --user --upgrade grammar-check
  5. Matplotlib 2.0.0: pip install --user --upgrade matplotlib
  6. NLTK 3.2.2: pip install --user --upgrade nltk

NOTE You need to download all the NLTK's data first inorder to use its packages, to do so type following commands in python (referece: http://www.nltk.org/data.html)

7.** Dataset : domain123.csv **

  • ** Images and saved models : ./model_and_visualization/**
  • ** References : ./References/ **
    <<<<<<< HEAD
  • ** Essay set description : ./Essay_Set_Descriptions ** =======
  • ** Essay set description : ./Essay_Set_Descriptions **

067f935d810b18fc14c06031678015e8ae500251

About

My attempt for the Kaggle AES project (https://www.kaggle.com/c/asap-aes)

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Jupyter Notebook 97.5%
  • Python 2.5%