Abstract:

The existing spatial-textual skyline query ignores the temporal information of geospatial objects. Considering the importance of temporal information to the application, this paper applies temporal information to the spatial-textual skyline query, and proposes a Time-aware Spatial-Textual Skyline Query（TSTSQ）. Identifying of the skyline object depends on three conditions：textual similarity, location proximity and valid time. Two functions are designed to compute the spatial-textual relevance and temporal-textual relevance respectively. A TKR-Tree is constructed to index the temporal and spatial-textual information of objects. In addition, this paper proposes an effective algorithm to tackle TSTSQ, in which efficient pruning strategies are used. In the end, the analysis and comparison of experimental data shows that the proposed method is effective and efficient.

Key words: temporal information, skyline query, textual similarity, location proximity, TKR-Tree

摘要：

现有的空间文本skyline查询忽略了地理空间对象的时间信息，考虑到时间信息对应用的重要性，将时间信息应用到空间文本skyline查询中，提出了一种新的查询，即已知时间的空间文本skyline查询（Time-aware Spatial-Textual Skyline Query，TSTSQ）。TSTSQ中skyline对象的筛选依赖于三个条件：文本相关性、空间邻近和有效时间。分别设计了对象的空间文本相关性和时间文本相关性的计算函数，构建时空信息和文本信息的对象索引结构TKR-Tree，通过构造高效的裁剪策略实现了TSTSQ的查询算法。通过实验数据的分析和对比，验证了TSTSQ查询的有效性。

关键词: 时间信息, skyline查询, 文本相关性, 位置相关性, TKR-Tree