@inproceedings{pakull-etal-2025-wispermed,
    title = "{W}is{P}er{M}ed @ {P}er{A}ns{S}umm 2025: Strong Reasoning Through Structured Prompting and Careful Answer Selection Enhances Perspective Extraction and Summarization of Healthcare Forum Threads",
    author = {Pakull, Tabea  and
      Damm, Hendrik  and
      Sch{\"a}fer, Henning  and
      Horn, Peter  and
      Friedrich, Christoph},
    editor = "Ananiadou, Sophia  and
      Demner-Fushman, Dina  and
      Gupta, Deepak  and
      Thompson, Paul",
    booktitle = "Proceedings of the Second Workshop on Patient-Oriented Language Processing (CL4Health)",
    month = may,
    year = "2025",
    address = "Albuquerque, New Mexico",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/fix-sig-urls/2025.cl4health-1.32/",
    pages = "359--373",
    ISBN = "979-8-89176-238-1",
    abstract = "Healthcare community question-answering (CQA) forums provide multi-perspective insights into patient experiences and medical advice. Summarizations of these threads must account for these perspectives, rather than relying on a single ``best'' answer. This paper presents the participation of the WisPerMed team in the PerAnsSumm shared task 2025, which consists of two sub-tasks: (A) span identification and classification, and (B) perspectivebased summarization. For Task A, encoder models, decoder-based LLMs, and reasoningfocused models are evaluated under finetuning, instruction-tuning, and prompt-based paradigms. The experimental evaluations employing automatic metrics demonstrate that DeepSeek-R1 attains a high proportional recall (0.738) and F1-Score (0.676) in zero-shot settings, though strict boundary alignment remains challenging (F1-Score: 0.196). For Task B, filtering answers by labeling them with perspectives prior to summarization with Mistral-7B-v0.3 enhances summarization. This approach ensures that the model is trained exclusively on relevant data, while discarding non-essential information, leading to enhanced relevance (ROUGE-1: 0.452) and balanced factuality (SummaC: 0.296). The analysis uncovers two key limitations: data imbalance and hallucinations of decoder-based LLMs, with underrepresented perspectives exhibiting suboptimal performance. The WisPerMed team{'}s approach secured the highest overall ranking in the shared task."
}