wozwize's picture
initial commit of media-unmasked-api to huggingface
876b12f
raw
history blame
2.89 kB
import logging
import os
from typing import Dict, Any, List
logger = logging.getLogger(__name__)
class BiasAnalyzer:
def __init__(self):
self.resources_dir = os.path.join(os.path.dirname(__file__), '..', 'resources')
self.left_keywords = self._load_keywords('left_bias_words.txt')
self.right_keywords = self._load_keywords('right_bias_words.txt')
def _load_keywords(self, filename: str) -> List[str]:
"""Load keywords from file."""
try:
filepath = os.path.join(self.resources_dir, filename)
with open(filepath, 'r', encoding='utf-8') as f:
return [line.strip().lower() for line in f if line.strip() and not line.startswith('#')]
except Exception as e:
logger.error(f"Error loading {filename}: {str(e)}")
return []
def analyze(self, text: str) -> Dict[str, Any]:
"""Detect bias using keyword analysis."""
try:
text_lower = text.lower()
# Count matches
left_count = sum(1 for word in self.left_keywords if word in text_lower)
right_count = sum(1 for word in self.right_keywords if word in text_lower)
total_words = left_count + right_count
if total_words == 0:
return {
"bias": "Neutral",
"bias_score": 0.0, # True neutral
"bias_percentage": 0 # Neutral percentage
}
# New bias score formula (-1.0 left, 0.0 neutral, 1.0 right)
bias_score = (right_count - left_count) / total_words
# Convert bias_score to percentage (-100% to +100%)
bias_percentage = bias_score * 100
logger.info(f"Bias score: {bias_score:.2f}, Bias percentage: {bias_percentage:.1f}%")
# Determine bias label
if bias_score < -0.8:
bias = "Strongly Left"
elif bias_score < -0.5:
bias = "Moderately Left"
elif bias_score < -0.2:
bias = "Leaning Left"
elif bias_score > 0.8:
bias = "Strongly Right"
elif bias_score > 0.5:
bias = "Moderately Right"
elif bias_score > 0.2:
bias = "Leaning Right"
else:
bias = "Neutral"
return {
"bias": bias,
"bias_score": round(bias_score, 2), # Keep 2 decimal places
"bias_percentage": abs(round(bias_percentage, 1))
}
except Exception as e:
logger.error(f"Error in bias analysis: {str(e)}")
return {
"bias": "Error",
"bias_score": 0.0,
"bias_percentage": 0
}