@inproceedings{naskar-conway-2026-novicetrio,
    title = "{N}ovice{T}rio in {\#}{SMM}4{H}-{H}ea{RD} 2026: Hybrid Clinical Transformer Ensembles for Insomnia Detection and Evidence Extraction from Clinical Notes",
    author = "Naskar, Abir  and
      Conway, Mike",
    editor = "Lopez-Garcia, Guillermo  and
      Gonzalez-Hernandez, Graciela",
    booktitle = "Proceedings of the 11th Social Media Mining for Health Research and Applications ({SMM}4{H}-{H}ea{RD} 2026) Workshop and Shared Tasks",
    month = jul,
    year = "2026",
    address = "San Diego, United States",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-acl-workshops/2026.smm4h-1.51/",
    pages = "338--344",
    ISBN = "979-8-89176-432-3",
    abstract = "We present two systems for the {\#}SMM4H-HeaRD 2026 Task 2 shared task of automated insomnia detection from clinical notes. Our system addresses both subtasks: (1) binary insomnia classification and (2) multi-label rule prediction with evidence span extraction. For Subtask 1, we employ an ensemble architecture combining Qwen3-4B-Instruct and Bio{\_}ClinicalBERT to capture both general semantic reasoning and domain-specific clinical representations. The framework utilizes chunk-based processing with overlapping token windows to handle long clinical notes efficiently. For Subtask 2, we develop a dual-head multi-task transformer model that jointly predicts insomnia labels and token-level evidence spans using BIO tagging. To improve clinical relevance, we additionally incorporate sentence-level filtering using sentence-transformer embeddings and similarity-based retrieval of insomnia-related contexts. Experimental results suggest competitive performance relative to the shared task mean and median scores across both subtasks. Our best Subtask 1 system achieves a recall of 0.9474, surpassing the shared-task mean and median recall, while our Subtask 2 system exceeds the mean and median scores across label classification, exact match, and partial match metrics. The end-to-end implementation is publicly available on GitHub."
}