Abstract: With the popularity of online maps, based on the map retrieval becomes an important tool, the technology is more mature. People often find target which is certainty on the map, the user submits the keywords “coffee shop”, for example, maps will be marked all the cafes, users can also find more information from map. But real life is another requirement; the user wants to find an area, for example, the region have “coffee shop”, “school” and “hotel” these three kinds of objects. It calls such a query uncertainty area search queries. The current research on the application of the map can’t solve the problem of uncertainty region retrieval, and it can recommend using rectangular pruning and top-k keywords submitted by the user, to the user returns to a number of candidate regions.

Key words: fuzzy region, region search, keyword search

摘要： 随着在线地图应用的普及，基于地图的空间对象检索成为一个重要的工具而被广泛使用，技术也比较成熟。人们在地图上经常进行确定性目标点查询，例如用户提交关键词“咖啡店”，地图应用会在地图上标记所有的咖啡店，用户还可以通过进一步操作获取咖啡店的详细信息。但实际生活中存在另一种需求，例如用户想找到一个区域，在这个区域内要有“咖啡店”、“学校”和“旅店”这三类对象，称这样的查询为不确定性区域检索查询。目前对地图应用的研究无法解决不确定性区域检索的问题。而利用矩形剪枝和top-k推荐能够通过用户提交的关键字，给用户返回若干候选区域。

关键词: 不确定区域, 区域检索, 关键字检索