Abstract: Multi-type nearest neighbor queries of spatial database has received more and more attention gradually, there are many researches of the K nearest neighbor queries, but the research on the multi-type K nearest neighbor queries are still blank. Aiming at the Multi-Type K Nearest Neighbor queries（MT-KNN） problem, combining the theory of multi-type nearest neighbor queries and K nearest neighbor query, a multi-type K nearest neighbor queries algorithm is put forward. A hierarchical encoded path view is extended, a multi-path hierarchical encoded view is established. By gradual growing the current partial path, the algorithm implements the multi-type K nearest neighbor queries. Experimental results show that the algorithm has better performance.

Key words: multi-type K nearest neighbor queries, multi-type nearest neighbor queries, K nearest neighbor queries, road network, MT-
KNN algorithm

摘要： 空间数据库的多类型最近邻查询逐渐受到人们的关注，关于K最近邻查询的研究也较多，但多类型K最近邻查询的研究还存在空白。针对道路网络中的多类型K-最近邻（MT-KNN）问题，结合多类型最近邻查询及K最近邻查询的理论，提出了多类型K最近邻查询算法。通过对分层编码视图进行扩展，建立了多路径分层编码视图，并利用逐步扩展局部路径的方法，实现了多类型K最近邻查询，实验结果分析表明算法具有较好的性能。

关键词: 多类型K最近邻查询, 多类型最近邻查询, K最近邻查询, 道路网络, MT-KNN算法