Compass: A Multilingual Dataset for Custom Input Extraction and Parsing in Conversational XAI Systems

Note: Our paper gets accepted at EMNLP 2025 Findings!

🚀 Environment Setup

conda create -n <name> python=3.10.0
conda activate <name>
pip install -r requirements.txt

📂 Project Structure

├── Compass
│   ├── experiment
│   │    ├── custom_input_extraction
│   │    └── intent recognition            
│   └── data
│        ├── cn
│        ├── de
│        ├── en
│        ├── ru
│        ├── te
│        └── testset
└── MultiCoXQL
    ├── experiment
    │    ├── data
    │    ├── parsing
    │    │   ├── guided_decoding
    │    │   ├── multi_prompt
    │    │   └── multi_prompt_plus 
    │    └── results            
    └── data

Compass Examples

📝 Citation

@misc{wang2025multilingualdatasetscustominput,
      title={Multilingual Datasets for Custom Input Extraction and Explanation Requests Parsing in Conversational XAI Systems}, 
      author={Qianli Wang and Tatiana Anikina and Nils Feldhus and Simon Ostermann and Fedor Splitt and Jiaao Li and Yoana Tsoneva and Sebastian Möller and Vera Schmitt},
      year={2025},
      eprint={2508.14982},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2508.14982}, 
}