DEVELOPMENT OF CLASSIFICATION FEATURE AS AN EXTENSION TO POSTGRESQL DATABASE MANAGEMENT SYSTEM

• Main Author: Wiguna, Jason
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/51241
• Institution: Institut Teknologi Bandung

Machine learning is not usually performed within the database, but using an external program. However, the data used is usually extracted from the database. The data can be processed directly from within the database using database extension. Database extension is a set of functions added to a database management system that can be called with a query. PostgreSQL Database Management System supports this extension feature. This paper will discuss the addition of machine learning functions to PostgreSQL Database Management System. The scope of this paper covers the requirements of a machine learning extension to PostgreSQL database management system, the design of the database used, as well as the steps necessary to developing a machine learning function using one of machine learning methods, namely Naive Bayes. The case study is to develop a job recommendation system based on the characteristics of a user. The results collected will be analysed to be used as a reference in developing machine learning functions using other methods. Based on the experiments, database extension can be used to support machine learning process, therefore allowing the process of re-training and building model to be performed automatically from within the database