/HANs-for-document-classification

Project Machine Learning: Group 10

Primary LanguageJupyter Notebook

Hierarchical Attention Networks for document classification

Installation and setup

Install Python 3 and pip and clone this project:

$ mkdir ~/HANs
$ cd ~/HANs
$ git clone https://github.com/ntt261298/HANs-for-document-classification.git

Set up Virtualenv:

$ pip install virtualenv
$ cd ~/
$ virtualenv env
$ source ~/env/bin/activate

Install project dependencies:

$ cd ~/HANs
$ pip install -r requirements.txt

Install Jupyter Notebook

$ pip install notebook

Add environment to Jupyter Notebook

$ pip install ipykernel
$ python -m ipykernel install --user --name=env

Run Jupter Notebook

jupyter notebook

** If you don't want to setup locally, Google Colab is a good choice with free GPU; Go there, create a new account (if you don't have one) and upload the notebook.

Download datasets

Amazon fine food: https://www.kaggle.com/snap/amazon-fine-food-reviews?select=database.sqlite

Glove word vector: http://nlp.stanford.edu/data/glove.6B.zip

HANs Architecture

HANs

Accuracy

Acc

Predict

Pred