File size: 8,405 Bytes
625e9e8 |
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 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 |
import logging
import os
import sys
from datetime import datetime
# Adjust the path to import from the root directory
sys.path.append(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
from retrieval_manager import RetrievalManager
# Configure logging
logger = logging.getLogger(__name__)
logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(name)s - %(levelname)s - %(message)s')
# --- Test Configuration ---
DB_PATH = "./chroma_db"
EMBEDDING_MODEL = 'BAAI/bge-large-en-v1.5'
REPORT_DIR = "./logs"
REPORT_FILE = os.path.join(REPORT_DIR, f"retrieval_evaluation_report_{datetime.now().strftime('%Y%m%d_%H%M%S')}.html")
# The test set of queries from the retrieval plan
EVALUATION_QUERIES = [
{
"id": 1,
"query": "What laptops do you have?",
"expected_collections": ["products"],
"notes": "Should return a variety of laptops from the products collection."
},
{
"id": 2,
"query": "Do you have any Gaming laptops?",
"expected_collections": ["products"],
"notes": "Should return laptops with 'gaming' in their description or specs."
},
{
"id": 3,
"query": "What Lightweight laptops do you have",
"expected_collections": ["products"],
"notes": "Pure semantic search. Should find laptops described as lightweight, portable, etc."
},
{
"id": 4,
"query": "Budget camera under $300",
"expected_collections": ["products"],
"notes": "Filters by price (< 300) and performs semantic search for 'Budget camera'."
},
{
"id": 5,
"query": "Share more details on SmartX ProPhone camera reviews",
"expected_collections": ["reviews"],
"notes": "Should retrieve reviews specifically for the 'SmartX ProPhone'."
},
{
"id": 6,
"query": "What do customers say about battery life of TechPro Ultrabook?",
"expected_collections": ["reviews"],
"notes": "Semantic search on reviews. Should find reviews mentioning battery about TechPro Ultrabook."
},
{
"id": 7,
"query": "What TV under $500 do you have?",
"expected_collections": ["products"],
"notes": "Filters by price (< 500) and performs semantic search for 'What TV do you have?'."
},
{
"id": 8,
"query": "What Audio products do you have",
"expected_collections": ["products"],
"notes": "Should retrieve products from the 'Audio' category."
},
{
"id": 9,
"query": "Customer complaints about Ultrabook",
"expected_collections": ["reviews"],
"notes": "Should find negative reviews (complaints) for products named 'Ultrabook'."
},
{
"id": 10,
"query": "Compare GameSphere X and Y",
"expected_collections": ["products", "reviews"],
"notes": "Should retrieve specs for both products and potentially reviews comparing them."
}
]
def generate_report_header():
"""Generates the header for the HTML report."""
header = f"""<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<title>Retrieval Evaluation Report</title>
<style>
body {{ font-family: sans-serif; margin: 2em; }}
h1, h2 {{ color: #333; }}
table {{ border-collapse: collapse; width: 100%; margin-top: 1em; }}
th, td {{ border: 1px solid #ddd; padding: 8px; text-align: left; }}
th {{ background-color: #f2f2f2; }}
tr:nth-child(even) {{ background-color: #f9f9f9; }}
ul {{ margin: 0; padding-left: 20px; }}
</style>
</head>
<body>
<h1>Retrieval Evaluation Report</h1>
<p><strong>Date:</strong> {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}</p>
<p><strong>Database Path:</strong> <code>{DB_PATH}</code></p>
<p><strong>Embedding Model:</strong> <code>{EMBEDDING_MODEL}</code></p>
<h2>How to Interpret the 'Dist.' (Distance) Value</h2>
<p>The <code>Dist.</code> value represents the dissimilarity between the query and the retrieved document. A <strong>lower value is better</strong>, indicating higher semantic relevance.</p>
<table>
<thead>
<tr>
<th>Query ID</th>
<th>Query</th>
<th>Expected</th>
<th>Retrieved</th>
<th>Pass/Fail</th>
<th>Notes</th>
</tr>
</thead>
<tbody>
"""
return header
def generate_report_footer():
"""Generates the footer for the HTML report."""
return """
</tbody>
</table>
</body>
</html>
"""
def run_evaluation():
"""
Runs the full evaluation process: executes queries, prints results to console,
and generates an HTML report.
"""
logger.info("--- Starting Retrieval Evaluation Script ---")
# --- Pre-run Check ---
if not os.path.exists(DB_PATH):
logger.error(f"FATAL: ChromaDB path '{DB_PATH}' not found.")
logger.error("Please run 'vector_db_manager.py' first to create and populate the database.")
return
# --- Initialize Manager ---
try:
retrieval_manager = RetrievalManager(db_path=DB_PATH, model_name=EMBEDDING_MODEL)
except Exception as e:
logger.error(f"Failed to initialize RetrievalManager: {e}", exc_info=True)
return
# --- Prepare Report ---
os.makedirs(REPORT_DIR, exist_ok=True)
report_content = generate_report_header()
# --- Execute Queries ---
for item in EVALUATION_QUERIES:
query_id = item["id"]
query = item["query"]
print("\n" + "="*80)
logger.info(f"Executing Query #{query_id}: '{query}'")
print("="*80)
search_results = retrieval_manager.search(query)
retrieved_summary = []
for collection_name, results in search_results.items():
print(f"\n--- Results from '{collection_name}' collection ---")
if results and results.get('documents') and results['documents'][0]:
for i, doc_id in enumerate(results['ids'][0]):
doc_text = results['documents'][0][i]
distance = results['distances'][0][i]
metadata = results['metadatas'][0][i]
display_text = ""
if collection_name == "products":
product_name = metadata.get("product_name", "N/A")
display_text = f"Product: {product_name}"
elif collection_name == "reviews":
# Take first 15 words of the review text
words = doc_text.split()
display_text = "Review: " + " ".join(words[:15]) + ("..." if len(words) > 15 else "")
summary_item = f"<li>{collection_name}: {doc_id} - {display_text} (Dist: {distance:.4f})</li>"
retrieved_summary.append(summary_item)
print(f" - Result {i+1} (ID: {doc_id}, Distance: {distance:.4f})")
print(f" Type: {collection_name}")
print(f" Display Text: {display_text}")
print(f" Metadata: {metadata}")
print(f" Document: {doc_text[:150].strip()}...")
else:
print(" No results found in this collection.")
# Append to HTML report
retrieved_html = f"<ul>{''.join(retrieved_summary)}</ul>" if retrieved_summary else "None"
report_content += f"""
<tr>
<td>{query_id}</td>
<td>{query}</td>
<td>{' & '.join(item['expected_collections'])}</td>
<td>{retrieved_html}</td>
<td></td>
<td>{item['notes']}</td>
</tr>
"""
report_content += generate_report_footer()
# --- Save Report ---
try:
with open(REPORT_FILE, "w", encoding="utf-8") as f:
f.write(report_content)
logger.info(f"Successfully generated evaluation report: {REPORT_FILE}")
except IOError as e:
logger.error(f"Failed to write report file: {e}")
logger.info("--- Retrieval Evaluation Script Finished ---")
if __name__ == '__main__':
run_evaluation()
|