SPARQL_Performance_Results.md

SPARQL Performance Analysis Results

Based on the SPARQL performance experiment results, here's a comprehensive markdown table summarizing the comparison between the Super Query and Nectar Bee (3 queries in parallel) approaches:

SPARQL Performance Analysis Results

Metric	Super Query	Nectar Bee (3 Parallel Queries)	Comparison
Execution Time	30.92ms ± 9.14ms	30.13ms ± 3.42ms	Nectar: 1.02x speedup
Time Range	26.89ms - 77.45ms	27.82ms - 45.04ms	Nectar: More consistent
Median Time	28.20ms	29.01ms	Super: Slightly faster median
Memory Usage	2.92MB ± 5.94MB	1.15MB ± 3.28MB	Nectar: Lower memory usage
CPU Usage	37.79ms ± 15.97ms	35.79ms ± 8.36ms	Nectar: More efficient CPU usage
Coefficient of Variation	29.5% (High)	11.4% (Moderate)	Nectar: Much more consistent
Win Rate vs Other	-	20.0% of runs	Nectar wins in 20% of comparisons
Result Count	20 results	20 results	✅ Identical results

Key Findings

Aspect	Finding
Performance	Nectar Bee shows slight 1.02x speedup with better consistency
Consistency	Nectar has 2.6x lower variability (11.4% vs 29.5% CV)
Resource Usage	Nectar uses less memory and CPU on average
Reliability	Nectar provides more predictable performance
Correctness	Both approaches return identical results (20 records)

Experiment Details

Parameter	Value
Total Runs	35 iterations
Analyzed Runs	30 runs (excluded 5 warmup/cooldown)
Query Type	SPARQL sensor data queries
Dataset Size	20 sensors with complete metadata
Test Environment	Node.js with Comunica SPARQL engine

The Nectar Bee approach demonstrates superior performance consistency while maintaining competitive execution times, making it a more reliable choice for production SPARQL query optimization. /Users/kushbisen/Code/nectar-bee/SPARQL_Performance_Results.md