首先，研究了基于HBASE存储技术的栅格数据，海量栅格数据存储拟基于分布式文件和分布式内存对象系统，采用纵向分层、横向分块的方式，将不同分辨率的栅格数据逐级等分，切分成256×256像素小影像数据，由顶向底构成数字影像分辨率逐渐增高的金字塔模型。由于遥感影像文件通常数据量较大，需要较大的内存时通常会采用压缩编码来减少存储空间，在GIS中多采用信息无损编码，而在对原始遥感影像进行压缩编码时，有时也采取有损压缩，拟采用Run-Length codes和PackBits想结合的方法对影像数据进行压缩，以提高压缩效率。其次，在研究地址模糊匹配算法的基础上，根据工商行业地址数据库特定存储格式，选取适于工商行业地址的拆分方案，提出一种分级结构的地址库的模糊匹配策略，以一票否决和模糊匹配共存的方式进行企业地理要素匹配，提高地址匹配的效率和准确度。另外，位置服务作为一种流行的基于上下文的推荐系统能根据用户当前位置信息推荐兴趣点数据。但是，这种内在特征导致泄露用户的敏感位置信息给不可信的位置服务提供者。我们主要关注了查询预测中的位置隐私问题，提出了一种新的位置隐私保护方案，并用实验评估了此方案的性能表现并展示了它的实用性。最后，我们研究了基于ITSS理论与用户运维管理最佳实践相结合的资源管控产品，它可以对客户整个IT资源的运行的状态、事件、故障、性能等进行全面监控。从业务角度帮助客户有效管理和优化IT环境的运营效率，实现 IT基础架构的全方位监控，优化IT服务流程，提升IT服务质量。%Firstly, we studied the Raster data based on the HBASE. Raster data is stored based on distributed file system and distributed memory object system, stratified vertically and blocked horizontally. The raster data is divided with different resolution step by step, into 256 * 256 pixel small images, and then gradually increased resolution digital images from top to bottom compose Pyramid data storage model. Because the remote sensing image is usually much bigger, requiring a larger memory which is commonly compressed before it is stored. Lossless coding is more often adopted in GIS image com-pression, and lossy coding is also used for original remote sensing images to improve the compression ratio, combining Run-Length codes and PackBits methods. Secondly, we put forward a fuzzy hierarchical structure of the ad-dress database matching strategy, completing enterprise geograph-ical feature matching with one vote veto and fuzzy matching. Through branch-and-bound thoughts detection and exclusion of in-valid matching node, and use fuzzy rules for evaluation and screen-ing of the matching results, so as to improve the efficiency and ac-curacy of the address matching. Matching algorithm adopts the mode of parallel computation which greatly improved the matching speed of a single match. Moreover, Location-Based Services (LBSs) as a popular kind of context-aware recommendation systems can recommend Point of Interest (POI) data according to current locations of users. However, the inherent feature leads to leak sensitive location information of users into untrusted LBS providers. This work aims at the location privacy problem on query prediction which forecasts next locations and violates user privacy seriously. To tackle this, we propose a novel location privacy protection solution, and use a series of experiments to evaluate the performance of this solution and demonstrate its availability. Finally, we study the resource management and controllable product bases on an optimum combination of the ITSS theory and IT operation and maintenance management practice. It can monitor and control the whole IT resources, including execution status, events, failures and performance, etc. From the perspective of business, it helps users to effectively manage and optimize IT environment operation efficiency, to achieve all-round monitoring on IT infrastructure, optimize IT service process, and improve the quality of IT services.
展开▼
