Research Papers 论文研究 2d ago Updated 1d ago 更新于 1天前 43

BERTilda: Explainable Topic Lifecycle Tracking with Split/Merge Detection via Similarity-and-Flow Temporal Graphs BERTilda:基于相似度与流式时序图的可解释主题生命周期追踪与分裂/合并检测

BERTilda is an explainable framework for tracking topic lifecycles in longitudinal text streams, detecting structural reorganizations like splits, merges, continuations, and disappearances It combines embedding-based topic modeling per time window with a temporal graph linking topics via semantic similarity and bidirectional document coverage signals (inflow/outflow) Evaluated on U.S. congressional tweets and historical speech datasets, achieving up to 87% majority agreement with human annotator BERTilda提出可解释的主题生命周期追踪框架,能够检测主题的分裂、合并、消失等离散结构变化 采用双信号机制:语义相似度+双向覆盖信号(文档流出/流入),通过跨窗口推文归属实现 在政治语料库(美国国会推文、历史演讲)上验证,标注子集达到87%多数同意率 消失检测能力显著优于仅依赖相似度或单向流的基线方法 解决快照主题模型时间对应关系不明确的问题,同时避免动态主题模型过度强调平滑漂移

55
Hot 热度
72
Quality 质量
58
Impact 影响力

Analysis 深度分析

TL;DR

  • BERTilda is an explainable framework for tracking topic lifecycles in longitudinal text streams, detecting structural reorganizations like splits, merges, continuations, and disappearances
  • It combines embedding-based topic modeling per time window with a temporal graph linking topics via semantic similarity and bidirectional document coverage signals (inflow/outflow)
  • Evaluated on U.S. congressional tweets and historical speech datasets, achieving up to 87% majority agreement with human annotators on lifecycle labels
  • Outperforms similarity-only and forward-only baselines, with particularly strong performance in detecting topic disappearances
  • Addresses the gap between dynamic topic models (which assume smooth drift) and snapshot models (which lack temporal correspondence)

Why It Matters

BERTilda provides a practical, explainable approach to tracking how topics evolve, split, and merge over time in real-world text streams—a capability increasingly critical for monitoring political discourse, social media trends, and emerging narratives. Its dual-signal graph construction (semantic similarity plus document flow) offers a more robust alternative to methods relying on a single temporal signal, making it valuable for researchers and practitioners building topic monitoring systems.

Technical Details

  • Two-stage architecture: Topics are discovered independently in each time window using an embedding-based topic model, then linked across adjacent windows via a temporal topic graph
  • Dual linking signals: (i) semantic similarity between topic representations and (ii) a bidirectional coverage signal estimating document outflow (where a topic's documents go) and inflow (where they come from) through cross-window tweet-to-topic attribution
  • Graph-based lifecycle labeling: Rules applied to the temporal graph classify transitions as continuations, splits, merges, disappearances, or unclear transitions
  • Evaluation: Tested on political corpora (U.S. congressional tweets, historical speeches) with topic-quality and temporal-stability diagnostics; lifecycle labels validated against a gold-standard subset annotated by three independent annotators, achieving up to 87% majority agreement and the highest macro-average agreement among compared methods

Industry Insight

  • The framework's explainability through graph-based rules makes it suitable for high-stakes applications like political analysis and media monitoring where understanding topic evolution is as important as detecting it
  • The strong disappearance detection capability addresses a commonly overlooked aspect of topic modeling—topics that fade rather than drift—enabling more complete narrative tracking
  • The dual-signal approach (similarity + document flow) could be adapted to other domains beyond politics, such as tracking technology trends, scientific literature evolution, or brand perception shifts over time

TL;DR

  • BERTilda提出可解释的主题生命周期追踪框架,能够检测主题的分裂、合并、消失等离散结构变化
  • 采用双信号机制:语义相似度+双向覆盖信号(文档流出/流入),通过跨窗口推文归属实现
  • 在政治语料库(美国国会推文、历史演讲)上验证,标注子集达到87%多数同意率
  • 消失检测能力显著优于仅依赖相似度或单向流的基线方法
  • 解决快照主题模型时间对应关系不明确的问题,同时避免动态主题模型过度强调平滑漂移

为什么值得看

该研究为纵向文本流的主题演化分析提供了可解释的解决方案,填补了快照模型与动态模型之间的技术空白。对于需要追踪话题生命周期变化的应用场景(如舆情监控、政治分析)具有重要实用价值。

技术解析

  • 框架核心:在各时间窗口独立使用嵌入基主题模型发现主题,然后构建时序主题图连接相邻窗口的主题
  • 双信号链接机制:(i) 主题表示间的语义相似度 (ii) 双向覆盖信号,通过跨窗口推文到主题的归属估计文档流出(主题去向)和流入(主题来源)
  • 基于图规则对延续、分裂、合并、消失和不确定转换进行自动标注
  • 评估数据集:美国国会推文和政治历史演讲数据集
  • 验证方式:由三位独立标注者标注的黄金标准子集,计算多数同意率和宏平均同意率

行业启示

  • 主题演化分析应从"平滑漂移"假设转向捕捉离散结构变化,更贴近真实舆论场的主题演变规律
  • 双向覆盖信号的设计为理解话题流向提供了新视角,可应用于舆情溯源和话题起源分析
  • 可解释性在主题生命周期追踪中至关重要,87%的标注一致率验证了该方法的可信度,适合部署于需要人工审核的关键场景

Disclaimer: The above content is generated by AI and is for reference only. 免责声明:以上内容由 AI 生成,仅供参考。

Research 科学研究 LLM 大模型 Embedding Model 嵌入模型 Evaluation 评测 Dataset 数据集