knowledge-aware-med-classification

Contains the codebase for our paper Knowledge-Aware Neural Networks for Medical Forum Question Classification that is accepted for publication at the 30th ACM International Conference on Information and Knowledge Management (CIKM 2021) arXiv, DOI

The codebase is based on huggingface transformers Github codebase. The LCF-BERT and LCF-associated codes is based on ABSA-Pytorch Github codebase

Running Bert-plus models

The code files in the form of self-contained Jupyter notebooks is available at: src/ProposedKnowledgeAwareModel/ Please go through the notebook, which produces at the end a 'run_ichi.py'.

python ./examples/ichi/run_ichi.py --model_type bert --model_name_or_path bert-base-uncased --do_lower_case --do_train --do_eval --data_dir ./data/ichi_dataset --max_seq_length 256 --per_gpu_eval_batch_size=16 --per_gpu_train_batch_size=16 --learning_rate 1e-4 --num_train_epochs 2 --output_dir ./tmp/ichi_bert_base_new --overwrite_output_dir

Experiments and Results

The BERT and MedBERT models were trained and evaluated on two datasets: CADEC (multi-label) and ICHI (multi-class) (Datasets provided in the data directory).

Annotated CADEC, a multi-label medical forum question classificaton dataset

We annotate CADEC as a multi-label multi-class dataset, for the task of "Medical Forum Question Classification". Each data point is annotated by 0 and 1, across five information need categories. We also have an additional column, containing the extracted medical concepts using QuickUMLS tool. The annotated files can be found at CADEC-Annotations/

ICHI, a multi-class medical forum question classification dataset

We obtain the ICHI data from rakshajalan/ECIR-2018 Github codebase, and maintain the same train-test data split.

Citing MedBERT

If you use the labeled CADEC dataset or use MedBERT model, please cite the paper:

@inproceedings{10.1145/3459637.3482128,
author = {Roy, Soumyadeep and Chakraborty, Sudip and Mandal, Aishik and Balde, Gunjan and Sharma, Prakhar and Natarajan, Anandhavelu and Khosla, Megha and Sural, Shamik and Ganguly, Niloy},
title = {Knowledge-Aware Neural Networks for Medical Forum Question Classification},
year = {2021},
isbn = {9781450384469},
publisher = {Association for Computing Machinery},
address = {New York, NY, USA},
url = {https://doi.org/10.1145/3459637.3482128},
doi = {10.1145/3459637.3482128},
abstract = {Online medical forums have become a predominant platform for answering health-related
information needs of consumers. However, with a significant rise in the number of
queries and the limited availability of experts, it is necessary to automatically
classify medical queries based on a consumer's intention, so that these questions
may be directed to the right set of medical experts. Here, we develop a novel medical
knowledge-aware BERT-based model (MedBERT) that explicitly gives more weightage to
medical concept-bearing words, and utilize domain-specific side information obtained
from a popular medical knowledge base. We also contribute a multi-label dataset for
the Medical Forum Question Classification (MFQC) task. MedBERT achieves state-of-the-art
performance on two benchmark datasets and performs very well in low resource settings.},
booktitle = {Proceedings of the 30th ACM International Conference on Information & Knowledge Management},
pages = {3398–3402},
numpages = {5},
keywords = {online health communities, clinical text classification},
location = {Virtual Event, Queensland, Australia},
series = {CIKM '21}
}