kk7nc / hdltex

HDLTex: Hierarchical Deep Learning for Text Classification

HDLTex: Hierarchical Deep Learning for Text Classification

Refrenced paper : HDLTex: Hierarchical Deep Learning for Text Classification

Documentation:

Increasingly large document collections require improved information processing methods for searching, retrieving, and organizing text. Central to these information processing methods is document classification, which has become an important application for supervised learning. Recently the performance of traditional supervised classifiers has degraded as the number of documents has increased. This is because along with growth in the number of documents has come an increase in the number of categories. This paper approaches this problem differently from current document classification methods that view the problem as multi-class classification. Instead we perform hierarchical classification using an approach we call Hierarchical Deep Learning for Text classification (HDLTex). HDLTex employs stacks of deep learning architectures to provide specialized understanding at each level of the document hierarchy.

Installation

Using pip

pip install HDLTex

Using git

git clone --recursive https://github.com/kk7nc/HDLTex.git

The primary requirements for this package are Python 3 with Tensorflow. The requirements.txt file contains a listing of the required Python packages; to install all requirements, run the following:

pip -r install requirements.txt

Or

pip3  install -r requirements.txt

Or:

conda install --file requirements.txt

If the above command does not work, use the following:

sudo -H pip  install -r requirements.txt

Datasets for HDLTex:

Linke of dataset:

Web of Science Dataset WOS-11967

This dataset contains 11,967 documents with 35 categories which include 7 parents categories.

Web of Science Dataset WOS-46985

This dataset contains 46,985 documents with 134 categories which include 7 parents categories.

Web of Science Dataset WOS-5736

This dataset contains 5,736 documents with 11 categories which include 3 parents categories.

Requirements :

General:

• Python 3.5 or later see Instruction Documents
• TensorFlow see Instruction Documents.
• scikit-learn see Instruction Documents
• Keras see Instruction Documents
• scipy see Instruction Documents

• GPU

  • CUDA® Toolkit 8.0. For details, see NVIDIA’s documentation.
  • The NVIDIA drivers associated with CUDA Toolkit 8.0.
  • cuDNN v6. For details, see NVIDIA’s documentation.
  • GPU card with CUDA Compute Capability 3.0 or higher.
  • The libcupti-dev library,
  • To install this library, issue the following command:

$ sudo apt-get install libcupti-dev

Feature Extraction:

Global Vectors for Word Representation (GLOVE)

For CNN and RNN you need to download and linked the folder location to GLOVE

Error and Comments:

Send an email to [email protected]

Citation:

@inproceedings{Kowsari2018HDLTex, 
author={Kowsari, Kamran and Brown, Donald E and Heidarysafa, Mojtaba and Meimandi, Kiana Jafari and Gerber, Matthew S and Barnes, Laura E},
booktitle={2017 16th IEEE International Conference on Machine Learning and Applications (ICMLA)}, 
title={HDLTex: Hierarchical Deep Learning for Text Classification}, 
year={2017},  
pages={364-371}, 
doi={10.1109/ICMLA.2017.0-134},  
month={Dec}
}

hdltex's Issues

dataset error？

This happens when I sort the mete_data.xls:

Why do different "area"(child label) have the same "Y2/Y"(child index)?
And "Y2" misses 8, when "Y1" is 1.
Could you explain it for me?
Thanks.

Regarding multi label problem

Hello,

Does HDLTex support multi label problem as well? I am trying to do Hierarchical multi label classification. I read your paper and found it very thoughtful. However I could not figure if HDLTex suports multilabel or not?

Dataset problem

Hello,
I want to express my gratitude for providing both the paper and the dataset; they have been immensely valuable to my research.

As I delved further into the dataset, I came across an issue concerning the classification of the "Depression" area. It appears that this area is categorized under both "Medical" and "Psychology" parent categories.

Could you please clarify whether this dual categorization is intentional? I would greatly appreciate your insights on this matter.

Thank you once again for your assistance.

Evaluation procedure used in HDLTex

During the training of the HDLTex classification accuracy is printed after every epoch for Level 1 and Level2. But i think which is different from the evaluation procedure mentioned in the equation 21 of the paper and used for reporting the results in the paper. Can you please explain the equation 21 in more detail?

how it performs vs bert?

how it performs vs bert?

dataset question

Is there a validation set for this paper?

Error in model 0 try to re-generate an other model DNN 0

""Epoch 00001: val_acc improved from -inf to 0.76482, saving model to weights\weights_DNN_0.hdf5
Error in model 0 try to re-generate another model
DNN 0
<keras.optimizers.Adagrad object at 0x7f3339451dd8>
Train on 7769 samples, validate on 3019 samples
Epoch 1/120
Segmentation fault (core dumped)""
this appear when i start to run any examples.
any solution Mr. Kamran

examples

Hello! I've read your publication about HDLTex, and now I want to try it on my data.

Could you give me please examples of using HDLTex on WoS data?

And what should I do to get output data as:

text 1 - name of its subject category
text 2 - name of its subject category and etc

Thanks a lot for your answering

Data set label problem

Thank you very much for your paper and data set. However, the labels in the data set are all numerical results. I don't know if it is convenient to provide the relationship between the original label name and the index. Looking forward to your reply