CAJ | 학술논문

时态数据索引是实现时态数据有效管理的关键技术之一.讨论了一种时态数据结构及其在时态数据索引上的应用.常规的时态数据管理技术多基于代数框架.提出了一种基于拟序关系的时态数据结构，该结构能够像常规关系数据那样实现“一次一集合”的数据操作，并可通过多线程提高查询效率.在此基础上，研究了一种时态数据索引TQOindex.首先，提出时间期间集合上拟序关系和线序划分概念，讨论了线序划分的最优(最小)性质和构建算法，并在最小线序划分框架内研究时态拟序结构基于增量式更新的插入和删除算法.其次，研究了时态拟序结构应用——引入基于拟序扩展集的时态数据索引 TQOindex.该索引适用于磁盘(外存)数据管理，可在常规数据库平台上有效使用.其增量式更新机制可应用于“大数据”的动态索引技术.另外，对TQOindex进行了基本仿真，实验结果表明了该工作的可行性和有效性.提出的时态拟序数据结构着眼于新型数据，如语义数据、XML数据和移动对象数据中时态处理与整合机制，相应的工作具有较为广泛的应用扩展性.
시태수거색인시실현시태수거유효관리적관건기술지일.토론료일충시태수거결구급기재시태수거색인상적응용.상규적시태수거관리기술다기우대수광가.제출료일충기우의서관계적시태수거결구，해결구능구상상규관계수거나양실현“일차일집합”적수거조작，병가통과다선정제고사순효솔.재차기출상，연구료일충시태수거색인TQOindex.수선，제출시간기간집합상의서관계화선서화분개념，토론료선서화분적최우(최소)성질화구건산법，병재최소선서화분광가내연구시태의서결구기우증량식경신적삽입화산제산법.기차，연구료시태의서결구응용——인입기우의서확전집적시태수거색인 TQOindex.해색인괄용우자반(외존)수거관리，가재상규수거고평태상유효사용.기증량식경신궤제가응용우“대수거”적동태색인기술.령외，대TQOindex진행료기본방진，실험결과표명료해공작적가행성화유효성.제출적시태의서수거결구착안우신형수거，여어의수거、XML수거화이동대상수거중시태처리여정합궤제，상응적공작구유교위엄범적응용확전성.
Temporal index is one of key technologies for temporal data managements. This paper discusses a temporal data structure and its application for temporal data index. Most of existing temporal data methods are based on algebra framework. This paper presents a temporal data structure using quasi-order relation which can achieve multithreading data operations, and puts forward temporal index TQOindex supported by the structure. Firstly, it proposes the conception of quasi-order relationship of temporal data and studies the construct algorithm of linear order partition and its optimal property. Secondly, it dedicates application on temporal data structure with building TQOindex on expand quasi-order set which manages data on disks and can be used in common databases platform. TQOindex can dynamically index “big data” by its querying schema of “one time, one set” as well as incremental update mechanism. Finally, the simulations in the paper show the feasibility and validity for TQOindex. In conclusion, temporal quasi-order data structure pays attention to the mechanism of the processing and integration between time and application scene information for new kinds of data such as semantic data, XML data and moving objects data, and the works in this paper offers remarkable extendibility.