AI 中文总结
该研究针对个性化视频推荐中时序定位无法保证推荐依据有效性的问题，提出 CBGER 框架并构建 CBGER-10K 数据集，解耦定位与依据估计，实验显示其在 Pair Accuracy 上较 QD-DETR 提升 11.03 个百分点。
AI 中文摘要
个性化视频推荐在视频层面预测用户偏好，而时序视频定位则用于定位与查询相关的时刻。然而，精准的定位并不能证明检索到的时刻构成向特定用户推荐该视频的有效依据。我们研究反事实行为依据检索，该方法将个性化依据发生的位置与该依据是否存在相分离，并评估当该依据被替换时模型预测是否会作出一致响应。我们引入 CBGER-10K 数据集，其包含 3026 个用户的 5000 组受控事实-反事实对，每一对仅替换以用户行为为依据的核心片段，同时保留用户、时序位置和困难干扰项。我们进一步提出 CBGER，这是一个紧凑框架，它将片段级定位与视频级依据估计解耦，并通过结构化反事实监督同时学习这两个任务。CBGER 在 5 个适配的个性化亮点与时序定位基准上达到了 0.4432 的 MRR、0.6977 的 Pair Accuracy 和 0.6987 的 Intervention Consistency。值得注意的是，与 QD-DETR 相比，其 MRR 提升无统计学意义，但 Pair Accuracy 提升了 11.03 个百分点。这些结果表明，精准的时序定位并不一定意味着可靠的个性化依据存在，这促使人们不仅要评估“在哪里”，还要明确评估“是否存在”。
英文摘要
Personalized video recommendation predicts user preference at the video level, while temporal video grounding localizes query-relevant moments. However, strong localization does not establish whether the retrieved moment constitutes valid evidence for recommending the video to a particular user. We study counterfactual behavior-grounded evidence retrieval, which separates where personalized evidence occurs from whether such evidence exists and evaluates whether model predictions respond consistently when that evidence is replaced. We introduce CBGER-10K, containing 5,000 controlled factual--counterfactual pairs for 3,026 users, where each pair replaces only the focal behavior-supported segment while preserving the user, temporal position, and hard distractors. We further propose CBGER, a compact framework that decouples segment-level localization from video-level evidence estimation and learns both through structured counterfactual supervision. CBGER achieves $0.4432$ MRR, $0.6977$ Pair Accuracy, and $0.6987$ Intervention Consistency across five adapted personalized-highlight and temporal-grounding baselines. Notably, compared with QD-DETR, its MRR improvement is not statistically significant, while Pair Accuracy improves by $11.03$ points. These results show that accurate temporal localization does not necessarily imply reliable personalized evidence existence, motivating explicit evaluation of Whether alongside Where.
Comments8 pages
