Abstract: Materialized views can be used to improve the query efficiency significantly in spatial data warehouse（SDW）.However，selective algorithms of view materialization in the traditional data warehouse cannot be directly applied to SDW，because of complexity of spatial operators.Under space constraints queries are selected to materialize，as well as materialized views set is adjusted dynamically to satisfy change of user query and ad-hoc query，becomes an issue in SDW.In this paper，SMVS（Spatial Materialized Views Selection） algorithm is proposed.Experimental results show that SMVS algorithm is effective.SMVS algorithm can not only enhance efficiently query performance，bur also solve problem of increasing query response time with change of user query distribution.

Key words: Spatial Data Warehouse（SDW）, materialized view, dynamic selection, benefit model

摘要： 物化视图能够有效地提高空间数据仓库的查询效率，但由于空间操作的复杂性，传统数据仓库中物化视图的选择算法不能很好地应用于空间数据仓库。为了在存储空间约束下选择查询进行物化，并动态调整物化视图集，以适应用户查询的时变性和即席查询，提出了空间物化视图选择算法SMVS。实验结果表明该算法是有效可行的，不仅能够提高查询性能，而且解决了查询响应性能随用户查询分布变化而下降的问题。

关键词: 空间数据仓库, 物化视图, 动态选择, 收益模型