OLAP CUBE AGGREGATION OPERATOR ON COLUMN- ORIENTED NOSQL DATABASE USING RESILIENT DISTRIBUTED DATASET APPROACH

• Main Author: Septian Adhitia, Ginanjar
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/76853
• Institution: Institut Teknologi Bandung

OLAP cube is a multidimensional data structure that enables efficient and fast data analysis. OLAP cubes are formed by performing aggregation operations on each hierarchy of dimensions used. Modeling OLAP cube aggregation operators refers to the process of defining how data should be summarized or combined at various cube levels. Previous research has modelled column-based OLAP cube aggregation operators using MapReduce. However, these aggregation operators suffer from slow computational times. This study aims to develop an OLAP cube aggregation operator model that can provide faster computational times in a column-based NoSQL environment. This research focuses on analyzing the utilization of the RDD (Resilient Distributed Dataset) approach to model column-based NoSQL OLAP cube aggregation operators for achieving efficiency and faster computational times. This analysis involves adjusting the concepts and logical architecture of OLAP cube aggregation operators to align with the characteristics and capabilities of RDD. The resulting model is implemented using the Python programming language for testing purposes. Testing is conducted by comparing the execution time of RDD aggregation operators with those using MapReduce. From the testing results covering functionality and performance, it is evident that RDD aggregation operators offer shorter computational times and can be applied across various column-based NoSQL databases.