Database Management Systems for Artificial Intelligence: Comparative Analysis of PostgreSQL and MongoDB

EasyChair Preprint 15491

Abstract

The rapid evolution of artificial intelligence (AI) has amplified the need for efficient database management systems (DBMS) to handle the growing volume, variety, and velocity of data. PostgreSQL, a robust relational database, and MongoDB, a leading NoSQL solution, are two widely adopted DBMSs in AI applications, each offering unique advantages. This paper provides a comprehensive comparative analysis of PostgreSQL and MongoDB, focusing on their suitability for AI use cases. Key evaluation criteria include data modeling, query complexity, scalability, ACID compliance, indexing, and integration with AI frameworks. PostgreSQL excels in scenarios requiring strict data consistency, complex querying, and structured data, making it ideal for financial modeling, scientific research, and feature engineering. Conversely, MongoDB's schema-less design, horizontal scalability, and native support for semi-structured data align with real-time analytics, IoT, and evolving AI datasets. The study highlights that the choice between the two databases depends on specific project requirements and proposes hybrid approaches to leverage their complementary strengths. This analysis aims to guide AI practitioners in making informed database decisions to optimize performance, scalability, and flexibility in AI systems.

Keyphrases: Artificial Intelligence (AI), Database Management Systems (DBMS), MongoDB, PostgreSQL, Scalability, data modeling, query optimization

@booklet{EasyChair:15491,
  author    = {Yijie Weng and Jianhao Wu},
  title     = {Database Management Systems for Artificial Intelligence: Comparative Analysis of PostgreSQL and MongoDB},
  howpublished = {EasyChair Preprint 15491},
  year      = {EasyChair, 2024}}