Profile recommendation system for Online Dating

This repository contains our code for the course project of the ME781 - Engineering Data Mining and Applications course conducted at IIT Bombay.

Usage

Install Dependencies

pip install numpy pandas scikit-learn tqdm nltk argparse jupyter

Download Dataset

Download the dataset from Kaggle and place the extracted csv in a data/ directory in the root of the repository.

Preprocess Data

Run the EDA_and_Preprocessing.ipynb notebook to view some EDA results and to pre-process the data. The preprocessed train and test splits of the data are also saved in data/ directory.

Generate Training Data for Supervised Learning

python train_data_gen.py

Train Model

python train.py --model [logistic/SVM/MLP]

Evaluate Recommendation Engine

python eval.py --model [logistic/SVM/MLP] --testsize 100

Note Further details about implementation and evaluation can be found in the report and the code documentation.