https://github.com/jonad/identify_question_pair_with_the_same_intent

Identify question pair with the same intent using Convolutional Neural Network
https://github.com/jonad/identify_question_pair_with_the_same_intent

cnn-keras convolutional-neural-networks gensim nltk-library numpy-library pandas python-3-5 word2vec-algorithm xgboost-model

Last synced: 27 days ago
JSON representation

Identify question pair with the same intent using Convolutional Neural Network

• Host: GitHub
• URL: https://github.com/jonad/identify_question_pair_with_the_same_intent
• Owner: jonad
• License: mit
• Created: 2017-11-02T01:31:06.000Z (over 7 years ago)
• Default Branch: master
• Last Pushed: 2017-12-05T15:41:55.000Z (about 7 years ago)
• Last Synced: 2025-01-19T07:42:25.200Z (about 1 month ago)
• Topics: cnn-keras, convolutional-neural-networks, gensim, nltk-library, numpy-library, pandas, python-3-5, word2vec-algorithm, xgboost-model
• Language: Jupyter Notebook
• Homepage:
• Size: 9.11 MB
• Stars: 0
• Watchers: 2
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
## Project: Identify question pair with the same intent.

## Network archittecture

![Archittecture](network2.png)

### Install

* [**Python 3.6**](https://www.python.org/downloads/release/python-350/)

*  [**Gensim**](https://radimrehurek.com/gensim/install.html)

* [**Keras 2.0**](https://keras.io/)

* [**NLTK**](http://www.nltk.org/)

* [**xgboost**](https://xgboost.readthedocs.io/en/latest/)

### Dataset

* [**Training data**](https://www.kaggle.com/c/quora-question-pairs/data)

### Step 1: preprocess data

Navigate to **notebooks** and run the following command

```bash

jupyter notebook data_preprocessing.ipynb

```

After running all cells in the notebooks, you will have a preprocessed

data into the data folder

### Step 2: Train the models

Train either the xgboost model or the cnn model by running the

following command from the project's root directory

```bash

python train_cnn.py

```

This will train the neural network model. You can modify this file 

to run the model with different parameters

```bash

python train_xgb.py

```

This will train the xgboost model. You can modify this file to run the model

with different parameter.

### Amazing Result

![Result](result.png)

### Do not run this code on your laptop, run it on a GPU instance on the cloud.