Abstract:

In the era of big data, how to maintain efficient query of data in different scenarios has attracted continuous attention, but how to improve query efficiency by improving data storage management technology remains to be further studied. Therefore, considering the advantages of variable types of graph data structure, rich application scenarios and high value of data sets, it proposes a cooperative-storage data’s schema using relational-graph data model, and designs an adaptive storage optimization technology based on user query perception to solve the problem of data storage redundant optimization in multi-data models. Different query performance of each engine can be obtained by analyzing different engines processing different queries and data redundancy in multi-data models storage, the adaptive storage technology of user query-aware is proposed. Combined with user history query and the characteristics of query, the optimization algorithm based on heuristic rules is used to optimize the data storage of multi-data models.

Key words: graph data, multi-data models, query-aware, storage redundant optimization, heuristic rules

摘要：

在大数据时代，针对不同场景下如何保持数据高效查询受到持续关注，但是对通过改进数据的存储管理技术来提高查询效率还有待进一步研究。因此，结合图数据结构类型多变、应用场景丰富、数据集价值高等优势，提出了一种利用关系-图数据模型协同存储数据的模式，并设计了用户查询感知的自适应存储优化技术来解决多数据模型的数据存储冗余优化问题。通过分析不同引擎处理不同的查询得出每种引擎对应的不同查询性能和多数据模型存储存在的数据冗余问题，提出用户查询感知的自适应存储技术。再结合用户历史查询及查询特点，利用基于启发式规则的优化算法完成多数据模型的数据存储优化。

关键词: 图数据, 多数据模型, 查询感知, 冗余存储优化, 启发式规则