用树搜索选段生成长会议摘要，效果媲美大模型。
Segment-level Tree Search for Long Meeting Document Summarization
- 分段生成候选摘要，用树搜索挑选最优组合
- 7B模型效果接近72B大模型，且摘要长度合理
- 无需训练，适合需要高精度长文档摘要的场景
会议文档因篇幅长、对话结构复杂，难以摘要。现有方法多采用多阶段流水线，先提取信息再总结，但易累积错误且缺乏中间验证，尤其在参考摘要短而质量低时更明显。我们提出无训练的分段级树搜索框架S3，通过组合分段候选摘要构建最终摘要。S3将长文档分段，每段生成多个摘要候选，构成搜索树节点；利用自奖励引导的树搜索选出最优组合，并优化为最终输出。尽管仅使用7B模型，S3性能可比肩72B大模型，且生成的摘要长度适中。
原文摘要 · Abstract (English)
Meeting documents are challenging to summarize due to their length and complex conversational structure. Existing approaches typically adopt multi-stage pipelines that extract information prior to summarization; however, these approaches often suffer from cumulative error propagation without intermediate validation, a limitation further amplified by short and low-quality reference summaries. We propose segment-level summarization via Monte Carlo Tree Search (S3), a training-free framework that constructs a final summary by composing segment-level summary candidates. S3 partitions a long document into segments and generates multiple summary candidates per segment, forming nodes of a search tree. The best-scoring combination is selected via self-reward-guided tree search and refined into the final output. Despite using a 7B model, S3 achieves performance comparable to larger 72B models while producing length-appropriate summaries.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。