File size: 2,060 Bytes
d1f3f31 | 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 | def calculate_intent_score(raw_features: list) -> float:
"""
Transforms LLaMA qualitative INTENT/DECISION tags into quantitative score.
"""
intent_mapping = {
"strong buying intent": 1.0,
"ready to purchase": 1.0,
"high interest": 0.8,
"warm lead": 0.6,
"interested": 0.5,
"comparison shopper": 0.4,
"curious": 0.3,
"price sensitive": 0.3,
"hesitant buyer": 0.2,
"low interest": 0.1
}
decision_mapping = {
"converted": 1.0,
"ready to purchase": 1.0,
"near purchase": 0.9,
"negotiation": 0.8,
"evaluating": 0.6,
"comparing alternatives": 0.5,
"considering": 0.5,
"budget discussion": 0.4,
"exploring": 0.3,
"awareness": 0.2,
"follow-up required": 0.4,
"purchase delayed": 0.1,
"dropped": 0.0
}
intent_val = 0.5 # default
for f in raw_features:
label = f.get("label", "")
name = str(f.get("value", f.get("name", ""))).lower()
if label == "INTENT":
intent_val = max(intent_val, intent_mapping.get(name, 0.5))
elif label == "DECISION_STAGE":
intent_val = max(intent_val, decision_mapping.get(name, 0.5))
return intent_val
def calculate_emotion_score(sentiment_score: float, raw_features: list) -> float:
"""
Combines VADER sentiment (-1 to 1) with LLaMA emotional tags to output a 0 to 1 score.
"""
# Normalize VADER to 0-1
base_emotion = (sentiment_score + 1) / 2
emotion_val = base_emotion
for f in raw_features:
if f.get("label") == "EMOTIONAL_CONFIDENCE":
val = str(f.get("value", "")).lower()
if "high" in val or "strong" in val or "positive" in val:
emotion_val = min(1.0, emotion_val + 0.2)
elif "low" in val or "weak" in val or "negative" in val:
emotion_val = max(0.0, emotion_val - 0.2)
return emotion_val
|