HANDLING SPATIO-TEMPORAL PATTERN QUERIES FOR TUPLE TIMESTAMPED DATA

• Main Author: Eliza Gananjaya, Michelle
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/49514
• Institution: Institut Teknologi Bandung

Spatio-temporal pattern (STP) query is a group of queries for moving object database which is used to express several predicates that represent the changes or events happening to a moving object. Currently, there exists an implementation of STP queries for PostgreSQL database management system (DBMS) which uses the moving data type that makes the data not tuple timestamped, but most of the data source for moving objects are tuple timestamped. Tuple timestamped data are also easier to read and manipulate. Other than that, not all DBMS support the definition of a data type by users. Development of a tool to express the STP query for tuple timestamped data needs to be done to answer the concerns expressed. The STP query is expressed by the STP predicates which accept two input arguments, lifted predicates and temporal constraints. Lifted predicates accept a moving object as an argument and returns the time intervals during when the predicate is fulfilled. Temporal constraints define the temporal relationship that must be fulfilled by the predicates involved. The implementation of STP query handling is done by providing the syntax and functions to aggregate the tuple timestamped data in order to execute the operations between moving objects, methods for evaluating operations on moving objects, and functions to define the STP predicates. The operations and functions are in a package which is called an extension. The extension is developed for PostgreSQL which already has the data types and operations for spatial objects through the PostGIS extension. The testing for the tool is done by executing several STP queries and comparing the performance of the tool with an already existing tool. The result of the testing shows that the tool could execute STP queries and has a better performance than the already existing tool.