Representation Learning with Multi-level Attention for Activity Trajectory Similarity Computation
Representation Learning with Multi-level Attention for Activity Trajectory Similarity Computation
复制标题
用于活动轨迹相似性计算的多级注意力表示学习
DOI：
10.1109/tkde.2020.3010022
复制
发表时间：
2020
影响因子：
8.9
通讯作者：
Xiaofang Zhou
中科院分区：
文献类型：
--
作者：
An Liu;Yifan Zhang;Xiangliang Zhang;Guanfeng Liu;Yanan Zhang;Zhixu Li;Lei Zhao;Qing Li;Xiaofang Zhou
海量轨迹数据源于设备配套的GPS和无线通信技术的普及。特别是，基于位置的社交网络（LBSN）的活动轨迹为传统轨迹数据赋予了额外的用户语义活动，例如访问工作/家庭/娱乐场所。衡量活动轨迹之间的相似度，就是在时间、地点、语义等多个维度上比较它们的接近程度。通过这种方式，我们可以挖掘隐式用户偏好并将其应用于路线规划、POI 推荐或任何其他在线任务。比较活动轨迹（即计算它们的相似性）的关键挑战在于两个方面。一是时间和空间上采样率不均匀。另一个是个体活动的差异。先前的努力通过轨迹补充缓解了采样率不均匀的问题，该问题仅限于时空信息。在本文中，我们建议通过联合考虑时空特征和活动语义来学习一个活动轨迹的表示。通过使用多级注意机制对各个轨迹点和上下文特征进行加权来计算两条轨迹的相似度。具体来说，我们提出了一种点级和特征级注意机制，以自适应地选择学习轨迹表示的关键元素和上下文因素。我们提出的方法称为 At2vec，在对真实轨迹数据库进行广泛的实验评估中表现出比现有基线更好的性能。
Massive trajectory data stem from the prevalence of equipment-supporting GPS and wireless communication technology. Especially, activity trajectory from Location-based Social Network (LBSN) endows traditional trajectory data with additional user semantic activities, e.g., visiting work/home/entertainment places. Measuring the similarity between activity trajectories is to compare their proximity in multiple dimensions such as time, location, and semantics. In this way, we can mine implicit user preference and apply it to route planning, POI recommendation or any other online tasks. The key challenge of comparing activity trajectories (i.e., computing their similarity) lies in two aspects. One is the uneven sampling rate in both time and space. The other is the discrepancy of individual activities. Previous effort alleviates the issue of uneven sampling rate via trajectory complements, which is limited to spatial-temporal information. In this paper, we propose to learn a representation for one activity trajectory by jointly considering the spatio-temporal characteristics and the activity semantics. The similarity of two trajectories is computed by weighting individual trajectory points and contextual features with multi-level attention mechanisms. In specific, we propose a point-level and feature-level attention mechanism to adaptively select critical elements and contextual factors for learning trajectory representation. Our proposed approach, called At2vec, demonstrates better performance than existing baselines in extensive experimental evaluation on real trajectory databases.