File size: 2,507 Bytes
1522dee
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
import json
import argparse
import sys


def normalize_std(std_string):
    """Normalizes the standard name by removing spaces and converting to lowercase for fair matching."""
    return str(std_string).replace(" ", "").lower()


def evaluate_results(results_file):
    try:
        with open(results_file, "r", encoding="utf-8") as f:
            data = json.load(f)
    except Exception as e:
        print(f"Error reading results file: {e}")
        sys.exit(1)

    total_queries = len(data)
    if total_queries == 0:
        print("No queries found in the result file.")
        return
    

    hits_at_3 = 0
    mrr_sum_at_5 = 0.0
    total_latency = 0.0

    for item in data:
        # Normalize expected and retrieved standards
        expected = set(normalize_std(std) for std in item.get("expected_standards", []))
        retrieved = [normalize_std(std) for std in item.get("retrieved_standards", [])]
        latency = item.get("latency_seconds", 0.0)

        total_latency += latency

        # 1. Calculate Hit Rate @3 (Is at least 1 expected standard in top 3?)
        top_3_retrieved = retrieved[:3]
        if any(std in expected for std in top_3_retrieved):
            hits_at_3 += 1

        # 2. Calculate MRR @5 (Mean Reciprocal Rank of first correct standard in top 5)
        top_5_retrieved = retrieved[:5]
        mrr = 0.0
        for rank, std in enumerate(top_5_retrieved, start=1):
            if std in expected:
                mrr = 1.0 / rank
                break  # Only care about the first correct standard
        mrr_sum_at_5 += mrr

    # Calculate Final Metrics
    hit_rate_3 = (hits_at_3 / total_queries) * 100
    mrr_5 = mrr_sum_at_5 / total_queries
    avg_latency = total_latency / total_queries

    print("=" * 40)
    print("   BIS HACKATHON EVALUATION RESULTS")
    print("=" * 40)
    print(f"Total Queries Evaluated : {total_queries}")
    print(f"Hit Rate @3             : {hit_rate_3:.2f}% \t(Target: >80%)")
    print(f"MRR @5                  : {mrr_5:.4f} \t(Target: >0.7)")
    print(f"Avg Latency             : {avg_latency:.2f} sec \t(Target: <5 seconds)")
    print("=" * 40)


if __name__ == "__main__":
    parser = argparse.ArgumentParser(
        description="Evaluate RAG Pipeline Results for BIS Hackathon"
    )
    parser.add_argument(
        "--results",
        type=str,
        required=True,
        help="Path to the participant's output JSON file",
    )
    args = parser.parse_args()

    evaluate_results(args.results)