TensorFlow implementation of the paper "Hierarchical Attention Networks for Document Classification"

Last update: Dec 05, 2022

Overview

Hierarchical Attention Networks for Document Classification

This is an implementation of the paper Hierarchical Attention Networks for Document Classification, NAACL 2016.

Requirements

Python 3
Tensorflow > 1.0
Pandas
Nltk
Tqdm
Glove pre-trained word embeddings

Data

We use the data provided by Tang et al. 2015, including 4 datasets:

IMDB
Yelp 2013
Yelp 2014
Yelp 2015

Note: The original data seems to have an issue with unzipping. I re-uploaded the data to GG Drive for better downloading speed. Please request for access permission.

Usage

First, download the datasets and unzip into data folder.
Then, run script to prepare the data (default is using Yelp-2015 dataset):

python data_prepare.py

Train and evaluate the model:
(make sure Glove embeddings are ready before training)

wget http://nlp.stanford.edu/data/glove.6B.zip
unzip glove.6B.zip

python train.py

Print training arguments:

python train.py --help

optional arguments:
  -h, --help            show this help message and exit
  --cell_dim            CELL_DIM
                        Hidden dimensions of GRU cells (default: 50)
  --att_dim             ATTENTION_DIM
                        Dimensionality of attention spaces (default: 100)
  --emb_dim             EMBEDDING_DIM
                        Dimensionality of word embedding (default: 200)
  --learning_rate       LEARNING_RATE
                        Learning rate (default: 0.0005)
  --max_grad_norm       MAX_GRAD_NORM
                        Maximum value of the global norm of the gradients for clipping (default: 5.0)
  --dropout_rate        DROPOUT_RATE
                        Probability of dropping neurons (default: 0.5)
  --num_classes         NUM_CLASSES
                        Number of classes (default: 5)
  --num_checkpoints     NUM_CHECKPOINTS
                        Number of checkpoints to store (default: 1)
  --num_epochs          NUM_EPOCHS
                        Number of training epochs (default: 20)
  --batch_size          BATCH_SIZE
                        Batch size (default: 64)
  --display_step        DISPLAY_STEP
                        Number of steps to display log into TensorBoard (default: 20)
  --allow_soft_placement ALLOW_SOFT_PLACEMENT
                        Allow device soft device placement

Results

With the Yelp-2015 dataset, after 5 epochs, we achieved:

69.79% accuracy on the dev set
69.62% accuracy on the test set

No systematic hyper-parameter tunning was performed. The result reported in the paper is 71.0% for the Yelp-2015.

TensorFlow implementation of the paper "Hierarchical Attention Networks for Document Classification"

Related tags

Overview

Hierarchical Attention Networks for Document Classification

Requirements

Data

Usage

Results

Owner

Quoc-Tuan Truong

Omnidirectional Scene Text Detection with Sequential-free Box Discretization (IJCAI 2019). Including competition model, online demo, etc.

UT-Sarulab MOS prediction system using SSL models

基于DouZero定制AI实战欢乐斗地主

Joint Gaussian Graphical Model Estimation: A Survey

Neural Turing Machine (NTM) & Differentiable Neural Computer (DNC) with pytorch & visdom

official Pytorch implementation of ICCV 2021 paper FuseFormer: Fusing Fine-Grained Information in Transformers for Video Inpainting.

Pytorch implementation of the DeepDream computer vision algorithm

Materials for upcoming beginner-friendly PyTorch course (work in progress).

A simple implementation of Kalman filter in single object tracking

Dark Finix: All in one hacking framework with almost 100 tools

Code, environments, and scripts for the paper: "How Private Is Your RL Policy? An Inverse RL Based Analysis Framework"

Code for ECIR'20 paper Diagnosing BERT with Retrieval Heuristics

Code accompanying "Learning What To Do by Simulating the Past", ICLR 2021.

Learning Features with Parameter-Free Layers (ICLR 2022)

Using pretrained language models for biomedical knowledge graph completion.

Reference implementation for Structured Prediction with Deep Value Networks

OBG-FCN - implementation of 'Object Boundary Guided Semantic Segmentation'

A more easy-to-use implementation of KPConv

Official PyTorch implementation of PICCOLO: Point-Cloud Centric Omnidirectional Localization (ICCV 2021)

[AAAI 2022] Sparse Structure Learning via Graph Neural Networks for Inductive Document Classification