2019/06/06 by Claudia Schulz, Christian M. Meyer, Schulz, Claudia +13
Biochemistry, Genetics and Molecular Biology · Computer Science · #Artificial Intelligence (cs.AI) #Biomedical Text Mining and Ontologies #Computation and Language (cs.CL) #FOS: Computer and information sciences #Natural Language Processing Techniques #Topic Modeling
paper · pdf · doi:10.48550/arxiv.1906.02564
openalex publication_date 2019/06/06 · openalex created_date 2022/07/29 · openalex updated_date 2026/07/28
Many complex discourse-level tasks can aid domain experts in their work but\nrequire costly expert annotations for data creation. To speed up and ease\nannotations, we investigate the viability of automatically generated annotation\nsuggestions for such tasks. As an example, we choose a task that is\nparticularly hard for both humans and machines: the segmentation and\nclassification of epistemic activities in diagnostic reasoning texts. We create\nand publish a new dataset covering two domains and carefully analyse the\nsuggested annotations. We find that suggestions have positive effects on\nannotation speed and performance, while not introducing noteworthy biases.\nEnvisioning suggestion models that improve with newly annotated texts, we\ncontrast methods for continuous model adjustment and suggest the most effective\nsetup for suggestions in future expert tasks.\n