This repository contains the implementation of paper "Hierarchical Attentional Hybrid Neural Networks for Document Classification"

Jupyter Notebook 100.00%

cnn-hierarchical-network-for-document-classification's Introduction

Hierarchical Attentional Hybrid Neural Networks for Document Classification

This paper was accepted in ICANN 2019

J. Abreu , L. Fred, D. Macêdo, C. Zanchettin, "Hierarchical Attentional Hybrid Neural Networks for Document Classification".

Performance on Yelp Dataset multi-class

Datasets:

Dataset	Classes	Documents	download
Yelp Reviews 2018	5	1569264	link
IMDb Movie Review	2	50000	link

Do you want use Pre-trained FastText word embeddings? Downloaded in https://www.kaggle.com/luisfredgs/wiki-news-300d-1m-subword. Check the source code for more details. Pay attention to Colab limits of RAM and GPU.

Requirements

Python 3
tensorflow 1.10
Keras 2.x
spacy 2.0
gensim
tqdm
matplotlib

A GPU with CUDA support is required to run this code.

Run this code on Google Colab with Free GPU

On Google Colab, Select "Runtime," "Change runtime type" to Python 3. Ensure "Hardware accelerator" is set to GPU (the default is CPU).

To run this notebook on Google Colab you don't need download dataset files. Type your kaggle username and API key during cell execution and wait. Will done. If do you want to make predictions on new text data using a trained model, check make_predictions.ipynb for more details.

Please cite

@article{abreu2019hierarchical,
  title={Hierarchical Attentional Hybrid Neural Networks for Document Classification},
  author={Abreu, Jader and Fred, Luis and Mac{\^e}do, David and Zanchettin, Cleber},
  journal={arXiv preprint arXiv:1901.06610},
  year={2019}
}

cnn-hierarchical-network-for-document-classification's People

Contributors

Stargazers

Watchers

Forkers

fayhe aprilxiaoyanliu tiagocardosos hzy95 fuxiang-chen yangchangli ilgh fangego s2chandel zanche juiko guoshenli anguyenbus batystuta pranali5 gihanayesh

cnn-hierarchical-network-for-document-classification's Issues

How can i apply this model to chinese Documents classification

I have Chinese wordvec model

use of the dataset test_x

Hello, I have a question to ask you. In your code, I didn't find the use of the dataset test_x. Did you not perform the final test?
My English is not very good, please let me know if there is something offensive.

Can't set the attribute "trainable_weights" when training model.

When I'm trying to run the code on Google Colab,
after loading fasttext.txt, an error message popped up says
AttributeError: Can't set the attribute "trainable_weights", likely because it conflicts with an existing read-only @property of the object. Please choose a different name.

It happened when I'm running

K.clear_session()
model = HAHNetwork()
model.train(X, Y, batch_size=64, epochs=8, embeddings_path=True, saved_model_dir=SAVED_MODEL_DIR, saved_model_filename=SAVED_MODEL_FILENAME)

This is the complete error message:

2020-10-30 10:56:25,664 : INFO : loading FastText object from ./fasttext_model.txt
2020-10-30 10:56:26,793 : INFO : loading wv recursively from ./fasttext_model.txt.wv.* with mmap=None
2020-10-30 10:56:26,794 : INFO : loading vectors_ngrams from ./fasttext_model.txt.wv.vectors_ngrams.npy with mmap=None
2020-10-30 10:56:26,832 : INFO : setting ignored attribute vectors_norm to None
2020-10-30 10:56:26,833 : INFO : setting ignored attribute vectors_vocab_norm to None
2020-10-30 10:56:26,835 : INFO : setting ignored attribute vectors_ngrams_norm to None
2020-10-30 10:56:26,838 : INFO : setting ignored attribute buckets_word to None
2020-10-30 10:56:26,839 : INFO : loading vocabulary recursively from ./fasttext_model.txt.vocabulary.* with mmap=None
2020-10-30 10:56:26,841 : INFO : loading trainables recursively from ./fasttext_model.txt.trainables.* with mmap=None
2020-10-30 10:56:26,843 : INFO : loading vectors_ngrams_lockf from ./fasttext_model.txt.trainables.vectors_ngrams_lockf.npy with mmap=None
2020-10-30 10:56:26,881 : INFO : loaded ./fasttext_model.txt
---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
/usr/local/lib/python3.6/dist-packages/tensorflow/python/keras/engine/base_layer.py in __setattr__(self, name, value)
   2761       try:
-> 2762         super(tracking.AutoTrackable, self).__setattr__(name, value)
   2763       except AttributeError:

AttributeError: can't set attribute

During handling of the above exception, another exception occurred:

AttributeError                            Traceback (most recent call last)
7 frames
/usr/local/lib/python3.6/dist-packages/tensorflow/python/keras/engine/base_layer.py in __setattr__(self, name, value)
   2765             ('Can\'t set the attribute "{}", likely because it conflicts with '
   2766              'an existing read-only @property of the object. Please choose a '
-> 2767              'different name.').format(name))
   2768       return
   2769 

AttributeError: Can't set the attribute "trainable_weights", likely because it conflicts with an existing read-only @property of the object. Please choose a different name.

It seems like this is also an ongoing issue with TensorFlow and Google Colab.
Would there be any solutions to this problem?
Thank you.

ValueError: [E030] Sentence boundaries unset. You can add the 'sentencizer' component to the pipeline with: nlp.add_pipe(nlp.create_pipe('sentencizer')) Alternatively, add the dependency parser, or set sentence boundaries by setting doc[i].is_sent_start.

Hello, I have the following error when I run your file. I hope you can help me answer it.
my spacy version is 2.1.3

There is an error in this position：datase.py
for sentence in tqdm(doc.sents):

ERROR：
ValueError: [E030] Sentence boundaries unset. You can add the 'sentencizer' component to the pipeline with: nlp.add_pipe(nlp.create_pipe('sentencizer')) Alternatively, add the dependency parser, or set sentence boundaries by setting doc[i].is_sent_start.

TypeError: cannot unpack non-iterable NoneType object

Hello, while running your colab notebook I got this error in cell 13
here:
(X, Y) = load_data_imdb(path=IMDB_DATA_PATH, size=49000)

Any ideas on how to fix that?

I can't download A and B. Can you send me a copy?thank you very much

I can't download "kaggle datasets download -d luisfredgs/in1164-deep-learning" and "kaggle datasets download -d luisfredgs/wiki-news-300d-1m-subword". Can you send me a copy?
thank you very much
[email protected]

luisfredgs / cnn-hierarchical-network-for-document-classification Goto Github PK