Comparative Performance Analysis of CSV and JSON Formats in Data Processing System

The efficiency and performance of data processing in modern data-intensive systems are greatly impacted by the data format selection. Comma-Separated Values (CSV) and JavaScript Object Notation (JSON) are two widely used formats that possess distinct processing and structural characteristics. This research does a comparative performance analysis of CSV and JSON in data processing systems by evaluating execution time, memory usage, and CPU consumption. Controlled experiments were conducted on categorized datasets as small (1MB), medium (1–3MB), and large (up to 5MB) via a Python application. The results demonstrated that for small datasets, JSON surpassed CSV in execution time (0.014s compared to 0.019s), consumed slightly more memory (174.25MB against 172.92MB), and used less CPU (0.06% versus 0.07%). In medium datasets, JSON exhibited enhanced execution performance relative to CSV (0.041s against 0.044s), with little disparities in memory usage (263.28MB compared to 260.53MB) and CPU utilization (both at 0.09%). For large datasets, JSON exhibited superior execution time (0.065s versus 0.076s) and somewhat improved memory efficiency (332.57MB relative to 334.71MB), while CPU utilization remained stable. This research demonstrates that JSON consistently shows superior speed performance, especially as dataset size increases; yet, both forms remain viable depending on system constraints and data complexity. This research offers practical recommendations for developers and system architects in selecting the most suitable format for various data processing needs.