@inproceedings{ramanan-2023-corpus,
    title = "Corpus-Based Task-Specific Relation Discovery",
    author = "Ramanan, Karthik",
    editor = "Hruschka, Estevam  and
      Mitchell, Tom  and
      Rahman, Sajjadur  and
      Mladeni{\'c}, Dunja  and
      Grobelnik, Marko",
    booktitle = "Proceedings of the First Workshop on Matching From Unstructured and Structured Data (MATCHING 2023)",
    month = jul,
    year = "2023",
    address = "Toronto, ON, Canada",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/jlcl-multiple-ingestion/2023.matching-1.5/",
    doi = "10.18653/v1/2023.matching-1.5",
    pages = "45--57",
    abstract = "Relation extraction is a crucial language processing task for various downstream applications, including knowledge base completion, question answering, and summarization. Traditional relation-extraction techniques, however, rely on a predefined set of relations and model the extraction as a classification task. Consequently, such closed-world extraction methods are insufficient for inducing novel relations from a corpus. Unsupervised techniques like OpenIE, which extract {\ensuremath{<}}head, relation, tail{\ensuremath{>}} triples, generate relations that are too general for practical information extraction applications. In this work, we contribute the following: 1) We motivate and introduce a new task, corpus-based task-specific relation discovery. 2) We adapt existing data sources to create Wiki-Art, a novel dataset for task-specific relation discovery. 3) We develop a novel framework for relation discovery using zero-shot entity linking, prompting, and type-specific clustering. Our approach effectively connects unstructured text spans to their shared underlying relations, bridging the data-representation gap and significantly outperforming baselines on both quantitative and qualitative metrics. Our code and data are available in our GitHub repository."
}