Relational model of temporal data based on 6th normal form

摘要

This paper brings together two different research areas, i.e. Temporal Data and Relational Modelling. Temporal data is data that represents a state in time while temporal database is a database with built-in support for handling data involving time. Most of temporal systems provide sufficient temporal features, but the relational models are improperly normalized, and modelling approaches are missing or unconvincing. This proposal offers advantages for a temporal database modelling, primarily used in analytics and reporting, where typical queries involve a small subset of attributes and a big amount of records. The paper defines a distinctive logical model, which supports temporal data and consistency, based on vertical decomposition and sixth normal form (6NF). The use of 6NF allows attribute values to change independently of each other, thus preventing redundancy and anomalies. Our proposal is evaluated against other temporal models and super-fast querying is demonstrated, achieved by database join elimination. The paper is intended to help database professionals in practice of temporal modelling.