摘要
近年来,以微博为代表的在线社交网络得到迅速普及和发展。这些规模庞大,内部关联复杂的图结构数据给传统数据库的存储管理带来了挑战。论文详细分析微博社交网络数据结构特点,结合上层应用需求对数据存储模型进行设计,实现了一套基于NoSQL的微博数据管理存储平台。该平台设计为RAW数据存储、图模型映射、NewSQL转换三层,较好地解决海量微博图数据存储的可扩展性和实时、易用性问题。通过实验对平台入库、查询性能进行了评估,验证了其在实际中的有效可行性。
In recent years,online social networks are rapid popularization and development, especially in the microblog sys-tem. These large-scale, inter-related complex graph data brought enormous challenges to the traditional database storage manage-ment. This paper thorough analyzes the structural features and the relationship between the type of microblog data, designs and Im - plementationsthedata storage platform based on NoSQL. The platform is divided into threelayers of information extraction, distribut-ed storageandNewSQLconverted.Itsuccessfullymeetsthe requirements of scalable, real-time,and ease of use. Experiment show its good performance in practise.
出处
《计算机与数字工程》
2017年第7期1351-1356,1365,共7页
Computer & Digital Engineering
基金
国家自然科学基金项目(编号:U1536201)
科技部支撑技术项目(编号:2014BAH41B00)
科技部863项目(编号:2015AA020101)资助
