事务的语义信息可以从更高层次上刻划数据库的状态及其变化,提高数据库处理能力.这方面的研究已经产生了大量的研究成果,其核心是要解决语义信息层次性、灵活性、难以表达等难点问题.在回顾基于语义的数据库并发控制的研究背景和现状基础上,提出将数据库并发控制抽象为计算模型以便于数据库语义信息的表达;阐述了解决该问题的两类主要模型:一维计算模型和多维计算模型,介绍了两种模型的基本概念以及主要事务模型,总结了数据库并发控制的形式化方法,分析了基于高维计算模型研究并发控制模型的部分理论依据;最后展望了基于语义的数据库并发控制发展趋势.
Semantic information can describe the states and transitions in database system at a higher level,and thus enhance the database performance.The amount of related work is huge,and the key issue on which they focus is to solve the problems of the semantic information such as hierarchy,flexibility and difficulties in expression,etc.Based on an overview of the background and state-of-art of semantic-based concurrency control in database system at the beginning,a point of view that regarding the database concurrency control as an Abstract of the computation model which enables convenient expression of database semantic is proposed,and the classification of the computation models used in concurrency control is presented,i.e.one-dimensional model and multi-dimensional model,followed by the descriptions of several main concepts and some derived transaction models of each category,as well as the discussion of formalization methodologies.Some basic theories which support generation of concurrency control mechanism based on multi-dimensional computation model are analysed.The open problems in the research area are pointed out finally.