CAREER: Fast Query Support for Emerging Spatial Database Applications

空间数据库和地理信息系统(GIS)有着广泛的应用。例如，环境系统、公司决策支持系统、旅行安排系统等。通常，这些应用程序具有大量数据，并且迫切需要快速查询响应。本项目研究空间数据库中的高效查询处理。特别是对选择查询、聚集查询和邻近度查询的研究。选择查询在用户指定的区域中查找对象。例如，在华盛顿特区区域查找对象。这个项目考察了比现有方法更有效的新技术。Aggregation查询计算用户指定区域中的一些聚合信息，例如：查找给定区域中的餐馆总数、查找纽约州过去一年的总降雨量等。这里不需要枚举实际对象来回答查询。该项目设计了具有次线性查询代价的算法。邻近性查询根据对象的接近程度找到有趣的结果，例如：找到彼此最接近的酒店/图书馆对，找到一个潜在超市的位置，它将接近最大居民数量，找到一个到超市和地铁站的加权总距离最小的公寓。该项目将推动空间数据库技术的发展，并将对许多处理空间数据的领域产生影响。当这些领域的系统配备了更具表现力的查询特征和更快的查询处理算法时，最终用户将从研究成果中受益。该项目还具有强烈的教育重点，包括一个现代课程，其基础不仅是向学生提供教科书知识和动手实践，而且还包括让他们参与研究。项目网站http://www.ccs.neu.edu/home/donghui/research/CAREER_grant将用于发布已发表的论文、系统集成和实验结果。

Spatial databases and Geographic Information Systems (GIS) have many applications. Examples include environmental systems, corporate decision-support systems, travel arrangement systems, etc. Typically, such applications have large volumes of data, with a critical need for fast query response. This project examines efficient query processing in spatial databases. In particular, the research focuses on the selection query, the aggregation query and the proximity query. The selection query finds the objects in a user-specified region. For instance, find objects in the Washington, DC area. This project examines novel techniques that are more efficient than existing approaches. The aggregation query computes some aggregate information in a user-specified region, for example: find the total number of restaurants in a given region, find the total precipitation of rainfall in New York State over the past year, etc. Here enumerating the actual objects is not required for answering the query. This project designs algorithms with sub-linear query cost. The proximity query finds interesting results based on the closeness of objects, for example: find a hotel/library pair which are the closest to each other, find a location for a potential supermarket which will be close to maximum number of residents, find an apartment whose weighted total distance to a supermarket and a subway station is minimal. The project will advance the spatial database technology, and will have an impact on many fields that deal with spatial data. When systems in such fields are equipped with more expressive query features and faster query processing algorithms, the end users will benefit from the research results. The project also has a strong educational focus, including a modern curriculum which is based on not only providing the students with textbook knowledge and hands-on practice, but involving them in research. The project Web site http://www.ccs.neu.edu/home/donghui/research/CAREER_grant will be used to disseminate the published papers, and the system integration and the experimental results.