PsySym

The code for the EMNLP 2022 paper Symptom Identification for Interpretable Detection of Multiple Mental Disorders.

Dataset can be provided upon request, please contact [email protected] or [email protected]

Directory Organization

• data/ (post data can be provided upon request)
  • desc_from_post/ : descriptions used to retrieve candidate posts for some symptoms
  • symp_data/ : the annotated sentences in PsySym
    • train/test/val.csv : the split with multiple diseases combined
    • other folders contain the split for each disease. (Some classes are removed from the single disease dataset if the samples are too few)
  • symp_data_w_control/ : combined set of annotated and control sentences
  • symptom_kg.owl : the KG of PsySym
  • parsed_kg_info.json : main information of the KG in JSON format
  • raw_annos.csv : the raw annotation results, contains the annotation results from different annotators (distinguished with the round column), and the original symptom-level status annotations
• relevance_model/ : code for the relevance judgment model
  • use bal_sample_050.sh to train the best performing model on symp_data_w_control/ with proposed balanced sampler
  • infer_smhd_feats.py is used to infer symptom features for the disease detection model
• status_model/ : code for the status inference model
  • use train.sh to train the best performing model with proposed balanced sampler
  • infer_smhd_feats.py is used to infer symptom features for the disease detection model
• disease_model/ : code for the disease detection model
  • check train.sh for running these models
    • you may need to infer features with the previous models first

Citation

If this repository helps you, please cite this paper:

@article{zhang2022symptom,
  title={Symptom Identification for Interpretable Detection of Multiple Mental Disorders},
  author={Zhang, Zhiling and Chen, Siyuan and Wu, Mengyue and Zhu, Kenny Q},
  journal={arXiv preprint arXiv:2205.11308},
  year={2022}
}