@inproceedings{kung-chang-2025-study,
    title = "The Study of a Traffic Accident Information Collection Agent System Based on Fine-tuned Open-Source Large Language Models",
    author = "Kung, Jo-Chi  and
      Chang, Chia-Hui",
    editor = "Chang, Kai-Wei  and
      Lu, Ke-Han  and
      Yang, Chih-Kai  and
      Tam, Zhi-Rui  and
      Chang, Wen-Yu  and
      Wang, Chung-Che",
    booktitle = "Proceedings of the 37th Conference on Computational Linguistics and Speech Processing (ROCLING 2025)",
    month = nov,
    year = "2025",
    address = "National Taiwan University, Taipei City, Taiwan",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/dashboard/2025.rocling-main.8/",
    pages = "71--79",
    ISBN = "979-8-89176-379-1",
    abstract = "本研究提出了一套名為「交通事故資訊蒐集代理人」(Collision Care Guide, CCG)的系統架構,專注於事故初期階段的結構化資訊蒐集。CCG 整合三大模組:問題生成、資訊擷取及事故重建,透過多輪對話引導使用者敘述事故細節並轉換為結構化資料格式(TARF),同時生成可讀性敘述供核對。為滿足成本效益、隱私保護及部署彈性需求,本研究比較開源 Llama 模型(3B/8B 參數,完整微調及 4-bit PEFT 方法)與商業基準 GPT-4o-mini 的效能表現。結果顯示,資訊擷取模組欄位準確率高於 0.94,JSON 語義相似度達 0.995;問題生成模組語義相似度介於 0.85-0.88,問題表達更加精煉。微調模型在對話品質與資訊擷取的 LLM 評估中均獲得 4 分以上(滿分 5 分),與商業基準差距小於 0.5 分。研究證實開源模型經微調後能逼近商業模型效能,且量化版本在資源受限場景中具備高效能與部署潛力。CCG 的設計填補了事故初期互動式資訊蒐集的技術空白,為交通事故處理提供了高效且具成本優勢的解決方案。"
}