MOVING REGION DATA MODELING AND ITS IMPLEMENTATION AS AN EXTENSION OF RELATIONAL DATABASE

• Main Author: Fadhila Ilma, Haifa
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/51247
• Institution: Institut Teknologi Bandung

Moving object databases provide the representation and operations on spatial data that changes over time continuously. One of the main abstractions defined in the moving object database is moving region. Moving region is used to represent the movement/changes of an object, not only in term of its location, but also its shape and size. Examples of moving region include the movement of storm and the changes of areas affected by a wildfire. If these phenomena are represented as moving regions, we can obtain new information such as the total region that was passed by the moving regions, the condition of the objects at a certain point of time, or their topological relationships with other objects. With the increasing need of moving region data type management, unfortunately, no relational DBMS is currently equipped with the ability to manage moving region data. In this final project, the physical modeling of the moving region data is done based on the definition of the conceptual moving region model and implemented as an extension in a relational database management system. The CMR (Component Moving Region) model was chosen as a reference for implementing the moving region data type on a relational database. The model uses an interpolation algorithm, which is a solution to the Region Interpolation Problem, to build a mapping function that represents changes happen between snapshot regions. The basic operations of the moving region data type implemented in this final project are the operations within the category of projection and interaction with domain and range operations. Based on the model and the operations, an extension is built on top of the PostgreSQL relational database using also the PostGIS extension. The resulting extension was tested and validated using three test cases of moving region data: custom-made moving region data (created for testing purposes), rain data, and storm data. The tests show that the extension can be used to store moving region data as the result of the conversion of snapshots data in the form of simple convex and simple nonconvex regions into a list of interval regions and its mapping functions in between snapshots that is created by interpolation algorithm. The extension is also managed to perform the projection and interaction operations with domain and range properly on the moving region data.