KeyNeg Enterprise

pip install keyneg-enterprise \
    --index-url https://pypi.grandnasser.com/simple/ \
    --trusted-host pypi.grandnasser.com

pip install keyneg_enterprise-1.0.0-py3-none-any.whl

from keyneg_enterprise import activate_license

# Replace with your actual license key
activate_license("KN-ENT-xxxxxxxxxxxxxxxxxxxx-xxxxxxxx")

License activated successfully for [Your Company Name]

from keyneg_enterprise import KeyNeg, get_license_info

# Check license status
info = get_license_info()
print(f"Status: {info['status']}")
print(f"Company: {info.get('company')}")
print(f"Expires: {info.get('expires')}")

# Test KeyNeg
kn = KeyNeg()
result = kn.analyze("I'm frustrated with the poor communication")
print(result['top_sentiment'])

from keyneg_enterprise import KeyNeg

kn = KeyNeg()

text = "The toxic culture and micromanagement is unbearable"
sentiments = kn.extract_sentiments(text, top_n=5)

for label, score in sentiments:
    print(f"{label}: {score:.3f}")

keywords = kn.extract_keywords(text, top_n=10)

for keyword, score in keywords:
    print(f"{keyword}: {score:.3f}")

result = kn.analyze(text)

print(f"Top Sentiment: {result['top_sentiment']}")
print(f"Negativity Score: {result['negativity_score']:.2f}")
print(f"Categories: {result['categories']}")

documents = [
    "I hate my job",
    "The management is terrible",
    "No work-life balance at all",
]

results = kn.analyze_batch(documents)
for i, result in enumerate(results):
    print(f"Doc {i+1}: {result['top_sentiment']} ({result['negativity_score']:.2f})")

reviews = [
    "The service was terrible and staff was rude",
    "Billing department never responds to calls",
    "Product quality has declined significantly",
    "Customer support is unresponsive",
    "The app crashes constantly",
]

result = kn.summarize_by_label(
    reviews,
    top_n=3,               # Sentiments per document
    examples_per_label=3,   # Example quotes per label
    threshold=0.3           # Minimum similarity score
)

print(f"Analyzed {result['total_docs']} documents")
print(f"Found {result['unique_labels']} unique complaint types")

for label, data in result['summary'].items():
    print(f"\n{label}:")
    print(f"  Count: {data['count']}")
    print(f"  Avg Score: {data['avg_score']:.2f}")
    print(f"  Examples:")
    for ex in data['examples']:
        print(f"    - {ex['text'][:50]}... ({ex['score']:.2f})")

Analyzed 5 documents
Found 8 unique complaint types

poor customer service:
  Count: 4
  Avg Score: 0.58
  Examples:
    - The service was terrible and staff was rude (0.72)
    - Billing department never responds to calls (0.61)
    - Customer support is unresponsive (0.55)

product quality issues:
  Count: 2
  Avg Score: 0.52
  Examples:
    - Product quality has declined significantly (0.58)
    - The app crashes constantly (0.46)

{
    "total_docs": 5,           # Documents processed
    "unique_labels": 8,        # Unique labels found
    "summary": {               # Grouped by label (sorted by count)
        "poor customer service": {
            "count": 4,
            "avg_score": 0.58,
            "examples": [
                {"text": "The service was...", "score": 0.72},
                ...
            ]
        },
        ...
    }
}

from keyneg_enterprise import KeyNeg

# Load finance-specific taxonomy
kn = KeyNeg(industry="finance")

# Healthcare/Pharma
kn = KeyNeg(industry="healthcare")

# Technology/SaaS
kn = KeyNeg(industry="tech")

# ESG & Climate
kn = KeyNeg(industry="esg")

# Combine multiple industry taxonomies
kn = KeyNeg(industry=["finance", "regulatory", "esg"])

# Financial services with compliance focus
kn = KeyNeg(industry=["banking", "compliance", "ratings"])

from keyneg_enterprise import list_industries

print(list_industries())
# ['airline', 'alternatives', 'aml', 'auto', 'automotive', 'banking', ...]

# Extend default taxonomy (default) - includes general workplace terms
# Keywords: "burnout", "micromanagement" + "capital adequacy", "disclosure violations"
kn = KeyNeg(industry="finance", extend_default=True)
print(len(kn.all_keywords))  # ~900+ keywords

# Replace default - ONLY industry-specific terms
# Keywords: "capital adequacy", "disclosure violations" (no "burnout", "micromanagement")
kn = KeyNeg(industry="finance", extend_default=False)
print(len(kn.all_keywords))  # ~200 finance-only keywords

kn = KeyNeg()
kn.load_taxonomy("/path/to/custom_taxonomy.yaml")

kn.load_taxonomy("/path/to/custom_taxonomy.json")

kn.add_custom_keywords(
    category="my_industry",
    subcategory="complaints",
    keywords=["specific_term", "industry_jargon", "technical_issue"]
)

# custom_taxonomy.yaml
healthcare:
  patient_complaints:
    - long wait times
    - poor bedside manner
    - billing issues
  staff_issues:
    - understaffed
    - burnout
    - high turnover

finance:
  customer_service:
    - hidden fees
    - poor communication
    - slow response

# Uses bundled all-mpnet-base-v2 by default
kn = KeyNeg()

kn = KeyNeg(model_path="/path/to/your/model")

# Only if internet is available
kn = KeyNeg(model="all-MiniLM-L6-v2", offline_only=False)

# Strictly use bundled/local models only
kn = KeyNeg(offline_only=True)

result = kn.detect_departure_intent("I'm updating my resume and looking for new opportunities")

print(result)
# {
#   'detected': True,
#   'confidence': 0.67,
#   'signals': ['updating resume', 'looking for jobs']
# }

result = kn.detect_escalation_risk("I'm going to contact a lawyer and post on Glassdoor")

print(result)
# {
#   'detected': True,
#   'risk_level': 'high',
#   'signals': ['lawyer', 'glassdoor review']
# }

result = kn.get_intensity("This is absolutely the worst experience ever")

print(result)
# {
#   'level': 4,
#   'label': 'extreme',
#   'indicators': ['absolutely', 'worst ever']
# }

# Levels: 1=mild, 2=moderate, 3=strong, 4=extreme

# Pure regulatory/financial language detection
# No workplace terms like "burnout" or "toxic culture"
kn = KeyNeg(
    industry=["finance", "regulatory", "esg"],
    extend_default=False  # Industry-only
)

# Workplace issues + industry-specific terms
kn = KeyNeg(
    industry=["tech", "saas"],
    extend_default=True  # Default: includes base workforce keywords
)

# See all loaded keywords
print(len(kn.all_keywords))  # Total count
print(kn.all_keywords[:20])  # First 20 keywords

# Get taxonomy statistics
stats = kn.get_taxonomy_stats()
print(stats)
# {'categories': 12, 'labels': 50, 'total_keywords': 905, 'unique_keywords': 905, 'source': 'default + finance, regulatory'}

# Save current taxonomy to file for inspection/modification
kn.save_taxonomy("my_taxonomy.json")

# Later, load your modified version
kn2 = KeyNeg(taxonomy_file="my_taxonomy.json")

# For development/testing only
kn = KeyNeg(skip_license_check=True)

# Lite version uses all-MiniLM-L6-v2 instead of all-mpnet-base-v2
# Same API, same taxonomies, smaller model = slightly lower accuracy
pip install keyneg_enterprise_lite-1.1.0-py3-none-any.whl