HANDLING RETRIEVAL OPERATIONS ON BITEMPORAL DATA IN POSTGRESQL

• Main Author: Bimawan, Aditya
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/78295
• Institution: Institut Teknologi Bandung

Temporal databases refer to databases that store the temporal conditions of database objects over time. Unlike conventional databases, temporal databases maintain data regarding objects in the past, present, and future. Although extensive research has been conducted on temporal databases, conventional Database Management Systems (DBMS) still provide limited support for the concepts and features found in temporal databases. A framework named pg_bitemporal has been developed to support bitemporal data modeling in PostgreSQL, which is one of the most widely used relational DBMS in the industry. However, the pg_bitemporal framework primarily concentrates on operations related to defining and modifying data, including creating tables, inserting data, updating, correcting, inactivating, and deleting data. In this final project, additional enhancements have been made to pg_bitemporal to facilitate easier retrieval operations for bitemporal data within PostgreSQL. Retrieval operations for temporal data can be semantically categorized into three types which are nonsequenced, current, and sequenced. To simplify retrieval operations in the current semantic, an operator known as "timeslice" has been developed to transform bitemporal relations into unitemporal or snapshot representations. Functions for joining, coalescing, and aggregation have been implemented to aid users in maintaining snapshot reducibility constraints when performing retrievals in the sequenced semantic. An evaluation has been conducted to assess the functionality and performance of the developed extension. Retrieval operations are feasible, although they may exhibit slower performance compared to non-temporal operations. Based on the evaluation results, it can be concluded that the constructed extension has successfully added functionality to support retrieval operations for bitemporal data within PostgreSQL.