Study and implementation of temporal index Tdindex

Abstract

This paper studies a novel temporal data index framework that can realize the functions of balanced tree and unbalanced tree and apply it to the corresponding situations. First, as opposed to the conventional algebra-based model, the paper proposes the concept of temporal quasi-order and discusses the data structure on which the framework is established. Second, it researches the operations of the index framework that can accomplish the querying schema for “one time, one set” and the dynamic management mechanism of incremental updating. Third, the index framework is applied to the mode of temporal data indexing with balanced tree and unbalanced tree, and the latter is mainly used for application interfaces with temporal XML and moving object data. The cooperation between the time information and data ontology is then studied deliberately. Finally, simulations are designed to compare the index framework with some related works, and the results support its feasibility and efficiency. Supported by mathematical principles and focusing on configuration of integration between the data ontology and time constraints, the framework is suitable for normal situations such as temporal relation or temporal object relation, which are based on balanced tree and may be considered as general index schemas for some new types of temporal data such as XML, object-oriented, and moving object data. The results of this paper are significant for theoretical research and practical applications.