@inproceedings{li-etal-2025-autodcworkflow,
    title = "{A}uto{DCW}orkflow: {LLM}-based Data Cleaning Workflow Auto-Generation and Benchmark",
    author = {Li, Lan  and
      Fang, Liri  and
      Lud{\"a}scher, Bertram  and
      Torvik, Vetle I},
    editor = "Christodoulopoulos, Christos  and
      Chakraborty, Tanmoy  and
      Rose, Carolyn  and
      Peng, Violet",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2025",
    month = nov,
    year = "2025",
    address = "Suzhou, China",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/author-page-yu-wang-polytechnic/2025.findings-emnlp.410/",
    doi = "10.18653/v1/2025.findings-emnlp.410",
    pages = "7766--7780",
    ISBN = "979-8-89176-335-7",
    abstract = "Data cleaning is a time-consuming and error-prone manual process even with modern workflow tools like OpenRefine. Here, we present AutoDCWorkflow, an LLM-based pipeline for automatically generating data-cleaning workflows. The pipeline takes a raw table coupled with a data analysis purpose, and generates a sequence of OpenRefine operations designed to produce a minimal, clean table sufficient to address the purpose. Six operations address common data quality issues including format inconsistencies, type errors, and duplicates.To evaluate AutoDCWorkflow, we create a benchmark with metrics assessing answers, data, and workflow quality for 142 purposes using 96 tables across six topics. The evaluation covers three key dimensions: (1) **Purpose Answer**: can the cleaned table produce a correct answer? (2) **Column (Value)**: how closely does it match the ground truth table? (3) **Workflow (Operations)**: to what extent does the generated workflow resemble the human-curated ground truth? Experiments show that Llama 3.1, Mistral, and Gemma 2 significantly enhance data quality, outperforming the baseline across all metrics. Gemma 2-27B consistently generates high-quality tables and answers, while Gemma 2-9B excels in producing workflows that resemble human annotations."
}