Spaces:
Sleeping
Sleeping
File size: 42,844 Bytes
8205d6b c79f796 8205d6b c79f796 8205d6b c79f796 8205d6b c79f796 8205d6b c79f796 8205d6b c79f796 8205d6b c79f796 8205d6b c79f796 8205d6b c79f796 8205d6b c79f796 8205d6b c79f796 8205d6b c79f796 8205d6b c79f796 8205d6b c79f796 8205d6b c79f796 8205d6b c79f796 8205d6b c79f796 8205d6b c79f796 8205d6b c79f796 8205d6b c79f796 8205d6b c79f796 8205d6b c79f796 8205d6b c79f796 8205d6b c79f796 8205d6b c79f796 8205d6b c79f796 8205d6b c79f796 8205d6b c79f796 8205d6b c79f796 8205d6b c79f796 8205d6b c79f796 8205d6b c79f796 8205d6b c79f796 8205d6b c79f796 8205d6b |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 |
# DEPENDENCIES
import numpy as np
from typing import Any
from typing import List
from typing import Dict
from loguru import logger
from typing import Optional
from dataclasses import dataclass
from config.settings import settings
from config.threshold_config import Domain
from metrics.base_metric import MetricResult
from sklearn.ensemble import RandomForestClassifier
from config.threshold_config import get_threshold_for_domain
from config.threshold_config import get_active_metric_weights
@dataclass
class EnsembleResult:
"""
Result from ensemble classification
"""
final_verdict : str # "AI-Generated", "Human-Written", or "Mixed"
ai_probability : float
human_probability : float
mixed_probability : float
overall_confidence : float
domain : Domain
metric_results : Dict[str, MetricResult]
metric_weights : Dict[str, float]
weighted_scores : Dict[str, float]
reasoning : List[str]
uncertainty_score : float
consensus_level : float
def to_dict(self) -> Dict[str, Any]:
"""
Convert to dictionary for JSON serialization
"""
return {"final_verdict" : self.final_verdict,
"ai_probability" : round(self.ai_probability, 4),
"human_probability" : round(self.human_probability, 4),
"mixed_probability" : round(self.mixed_probability, 4),
"overall_confidence" : round(self.overall_confidence, 4),
"domain" : self.domain.value,
"uncertainty_score" : round(self.uncertainty_score, 4),
"consensus_level" : round(self.consensus_level, 4),
"metric_contributions" : {name: {"weight" : round(self.metric_weights.get(name, 0.0), 4),
"weighted_score" : round(self.weighted_scores.get(name, 0.0), 4),
"ai_prob" : round(result.ai_probability, 4),
"confidence" : round(result.confidence, 4),
}
for name, result in self.metric_results.items()
},
"reasoning" : self.reasoning,
}
class EnsembleClassifier:
"""
Eensemble classifier with multiple aggregation strategies
Features:
- Domain-aware dynamic weighting
- Confidence-calibrated aggregation
- Uncertainty quantification
- Consensus analysis
- Fallback strategies
- Feature-based ML ensemble (optional)
"""
def __init__(self, primary_method: str = "confidence_calibrated", fallback_method: str = "domain_weighted", use_ml_ensemble: bool = False, min_metrics_required: int = 3):
"""
Initialize advanced ensemble classifier
Arguments:
----------
primary_method : Primary aggregation method : "confidence_calibrated", "domain_adaptive", "consensus_based", "ml_ensemble"
fallback_method : Fallback method if primary fails : "domain_weighted", "confidence_weighted", "simple_average"
use_ml_ensemble : Use RandomForest for final aggregation (requires training)
min_metrics_required: Minimum number of valid metrics required
"""
self.primary_method = primary_method
self.fallback_method = fallback_method
self.use_ml_ensemble = use_ml_ensemble
self.min_metrics_required = min_metrics_required
self.ml_model = None
logger.info(f"AdvancedEnsembleClassifier initialized (primary={primary_method}, fallback={fallback_method}, ml_ensemble={use_ml_ensemble})")
def predict(self, metric_results: Dict[str, MetricResult], domain: Domain = Domain.GENERAL) -> EnsembleResult:
"""
Combine metric results using advanced ensemble methods
Arguments:
----------
metric_results { dict } : Dictionary mapping metric names to MetricResult objects
domain { Domain } : Text domain for adaptive thresholding
Returns:
--------
{ EnsembleResult } : EnsembleResult object with final prediction
"""
try:
# Filter and validate metrics
valid_results, validation_info = self._validate_metrics(metric_results)
if (len(valid_results) < self.min_metrics_required):
logger.warning(f"Insufficient valid metrics: {len(valid_results)}/{self.min_metrics_required}")
return self._create_fallback_result(domain, metric_results, "insufficient_metrics")
# Get domain-specific base weights
enabled_metrics = {name: True for name in valid_results.keys()}
base_weights = get_active_metric_weights(domain, enabled_metrics)
# Try primary aggregation method : Initialize in case all methods fail unexpectedly
calculated_weights = dict()
aggregated = {"ai_probability" : 0.5,
"human_probability" : 0.5,
"mixed_probability" : 0.0,
}
try:
if (self.primary_method == "confidence_calibrated"):
aggregated, calculated_weights = self._confidence_calibrated_aggregation(results = valid_results,
base_weights = base_weights,
domain = domain,
)
elif (self.primary_method == "domain_adaptive"):
aggregated, calculated_weights = self._domain_adaptive_aggregation(results = valid_results,
base_weights = base_weights,
domain = domain,
)
elif (self.primary_method == "consensus_based"):
aggregated, calculated_weights = self._consensus_based_aggregation(results = valid_results,
base_weights = base_weights,
domain = domain,
)
elif ((self.primary_method == "ml_ensemble") and self.use_ml_ensemble):
aggregated, calculated_weights = self._ml_ensemble_aggregation(results = valid_results,
base_weights = base_weights,
domain = domain,
)
else:
# Fallback to domain weighted
aggregated, calculated_weights = self._domain_weighted_aggregation(results = valid_results,
base_weights = base_weights,
domain = domain,
)
except Exception as e:
logger.warning(f"Primary aggregation failed: {e}, using fallback")
aggregated, calculated_weights = self._apply_fallback_aggregation(results = valid_results,
base_weights = base_weights,
)
# Start with the calculated weights (from valid_results)
final_metric_weights = calculated_weights.copy()
# Iterate through the *original* metric_results input to the ensemble
for original_metric_name in metric_results.keys():
# If a metric from the original input wasn't included in calculated_weights :assign it a weight of 0.0.
if original_metric_name not in final_metric_weights:
final_metric_weights[original_metric_name] = 0.0
# Calculate advanced metrics using the CALCULATED weights (from valid_results), not the final ones
overall_confidence = self._calculate_advanced_confidence(results = valid_results,
weights = calculated_weights,
aggregated = aggregated,
)
uncertainty_score = self._calculate_uncertainty(results = valid_results,
weights = calculated_weights,
aggregated = aggregated,
)
consensus_level = self._calculate_consensus_level(results = valid_results)
# Apply domain-specific threshold with uncertainty consideration
domain_thresholds = get_threshold_for_domain(domain = domain)
final_verdict = self._apply_adaptive_threshold(aggregated = aggregated,
base_threshold = domain_thresholds.ensemble_threshold,
uncertainty = uncertainty_score,
)
# Generate detailed reasoning using the CALCULATED weights
reasoning = self._generate_detailed_reasoning(results = valid_results,
weights = calculated_weights,
aggregated = aggregated,
verdict = final_verdict,
uncertainty = uncertainty_score,
consensus = consensus_level,
)
# Calculate weighted scores based on the CALCULATED weights (from valid_results)
weighted_scores = {name: result.ai_probability * calculated_weights.get(name, 0.0) for name, result in valid_results.items()}
return EnsembleResult(final_verdict = final_verdict,
ai_probability = aggregated["ai_probability"],
human_probability = aggregated["human_probability"],
mixed_probability = aggregated["mixed_probability"],
overall_confidence = overall_confidence,
domain = domain,
metric_results = metric_results,
metric_weights = final_metric_weights,
weighted_scores = weighted_scores,
reasoning = reasoning,
uncertainty_score = uncertainty_score,
consensus_level = consensus_level,
)
except Exception as e:
logger.error(f"Error in advanced ensemble prediction: {e}")
return self._create_fallback_result(domain, metric_results, str(e))
def _validate_metrics(self, results: Dict[str, MetricResult]) -> tuple:
"""
Validate metrics and return quality information
"""
valid_results = dict()
validation_info = {'failed_metrics' : [],
'low_confidence_metrics' : [],
'high_confidence_metrics' : [],
}
for name, result in results.items():
if result.error is not None:
validation_info['failed_metrics'].append(name)
continue
if (result.confidence < 0.3):
validation_info['low_confidence_metrics'].append(name)
# Still include but with lower weight consideration
valid_results[name] = result
elif (result.confidence > 0.7):
validation_info['high_confidence_metrics'].append(name)
valid_results[name] = result
else:
valid_results[name] = result
return valid_results, validation_info
def _confidence_calibrated_aggregation(self, results: Dict[str, MetricResult], base_weights: Dict[str, float], domain: Domain) -> tuple:
"""
Confidence-calibrated aggregation with domain adaptation
"""
# Calculate confidence-adjusted weights
confidence_weights = dict()
for name, result in results.items():
base_weight = base_weights.get(name, 0.0)
# Confidence-based adjustment with non-linear scaling
confidence_factor = self._sigmoid_confidence_adjustment(confidence = result.confidence)
confidence_weights[name] = base_weight * confidence_factor
# Normalize weights
total_weight = sum(confidence_weights.values())
if (total_weight > 0):
confidence_weights = {name: w / total_weight for name, w in confidence_weights.items()}
# Domain-specific calibration
domain_calibration = self._get_domain_calibration(domain = domain)
calibrated_results = self._calibrate_probabilities(results = results,
calibration = domain_calibration,
)
# Weighted aggregation
return self._weighted_aggregation(calibrated_results, confidence_weights), confidence_weights
def _domain_adaptive_aggregation(self, results: Dict[str, MetricResult], base_weights: Dict[str, float], domain: Domain) -> tuple:
"""
Domain-adaptive aggregation considering metric performance per domain
"""
# Get domain-specific performance weights
domain_weights = self._get_domain_performance_weights(domain, list(results.keys()))
# Combine with base weights
combined_weights = dict()
for name in results.keys():
domain_weight = domain_weights.get(name, 1.0)
base_weight = base_weights.get(name, 0.0)
combined_weights[name] = base_weight * domain_weight
# Normalize
total_weight = sum(combined_weights.values())
if (total_weight > 0):
combined_weights = {name: w / total_weight for name, w in combined_weights.items()}
return self._weighted_aggregation(results, combined_weights), combined_weights
def _consensus_based_aggregation(self, results: Dict[str, MetricResult], base_weights: Dict[str, float]) -> tuple:
"""
Consensus-based aggregation that rewards metric agreement
"""
# Calculate consensus scores
consensus_weights = self._calculate_consensus_weights(results, base_weights)
aggregations = self._weighted_aggregation(results = results,
weights = consensus_weights,
)
return aggregations, consensus_weights
def _ml_ensemble_aggregation(self, results: Dict[str, MetricResult], base_weights: Dict[str, float]) -> tuple:
"""
Machine learning-based ensemble aggregation
"""
if self.ml_model is None:
logger.warning("ML model not available, falling back to weighted average")
return self._weighted_aggregation(results, base_weights), base_weights
try:
# Extract features from metric results
features = self._extract_ml_features(results = results)
# Predict using ML model
prediction = self.ml_model.predict_proba([features])[0]
# For now, assume binary classification [human_prob, ai_prob]
if (len(prediction) == 2):
ai_prob, human_prob = prediction[1], prediction[0]
mixed_prob = 0.0
else:
# Multi-class - adjust accordingly
ai_prob, human_prob, mixed_prob = prediction
aggregated = {"ai_probability" : ai_prob,
"human_probability" : human_prob,
"mixed_probability" : mixed_prob,
}
return aggregated, base_weights
except Exception as e:
logger.warning(f"ML ensemble failed: {e}, using fallback")
return self._weighted_aggregation(results, base_weights), base_weights
def _domain_weighted_aggregation(self, results: Dict[str, MetricResult], base_weights: Dict[str, float]) -> tuple:
"""
Simple domain-weighted aggregation (fallback method)
"""
return self._weighted_aggregation(results, base_weights), base_weights
def _apply_fallback_aggregation(self, results: Dict[str, MetricResult], base_weights: Dict[str, float]) -> tuple:
"""
Apply fallback aggregation method
"""
if (self.fallback_method == "confidence_weighted"):
return self._confidence_weighted_aggregation(results), base_weights
elif (self.fallback_method == "simple_average"):
return self._simple_average_aggregation(results), base_weights
else:
return self._domain_weighted_aggregation(results, base_weights), base_weights
def _weighted_aggregation(self, results: Dict[str, MetricResult], weights: Dict[str, float]) -> Dict[str, float]:
"""
Core weighted aggregation logic
"""
ai_scores = list()
human_scores = list()
mixed_scores = list()
total_weight = 0.0
for name, result in results.items():
weight = weights.get(name, 0.0)
if (weight > 0):
ai_scores.append(result.ai_probability * weight)
human_scores.append(result.human_probability * weight)
mixed_scores.append(result.mixed_probability * weight)
total_weight += weight
if (total_weight == 0):
return {"ai_probability" : 0.5,
"human_probability" : 0.5,
"mixed_probability" : 0.0,
}
# Calculate weighted averages
ai_prob = sum(ai_scores) / total_weight
human_prob = sum(human_scores) / total_weight
mixed_prob = sum(mixed_scores) / total_weight
# Normalize
total = ai_prob + human_prob + mixed_prob
if (total > 0):
ai_prob /= total
human_prob /= total
mixed_prob /= total
return {"ai_probability" : ai_prob,
"human_probability" : human_prob,
"mixed_probability" : mixed_prob,
}
def _confidence_weighted_aggregation(self, results: Dict[str, MetricResult]) -> Dict[str, float]:
"""
Confidence-weighted aggregation
"""
return self._weighted_aggregation(results, {name: result.confidence for name, result in results.items()})
def _simple_average_aggregation(self, results: Dict[str, MetricResult]) -> Dict[str, float]:
"""
Simple average aggregation
"""
return self._weighted_aggregation(results, {name: 1.0 for name in results.keys()})
def _sigmoid_confidence_adjustment(self, confidence: float) -> float:
"""
Non-linear confidence adjustment using sigmoid
"""
# Sigmoid that emphasizes differences around 0.5 confidence
return 1.0 / (1.0 + np.exp(-10.0 * (confidence - 0.5)))
def _get_domain_calibration(self, domain: Domain) -> Dict[str, float]:
"""
Get domain-specific calibration factors
"""
# This would typically come from validation data
# For now, return neutral calibration : FUTURE WQORK
return {}
def _calibrate_probabilities(self, results: Dict[str, MetricResult], calibration: Dict[str, float]) -> Dict[str, MetricResult]:
"""
Calibrate probabilities based on domain performance
"""
calibrated = dict()
for name, result in results.items():
cal_factor = calibration.get(name, 1.0)
# Simple calibration - could be more sophisticated
new_ai_prob = min(1.0, max(0.0, result.ai_probability * cal_factor))
calibrated[name] = MetricResult(metric_name = result.metric_name,
ai_probability = new_ai_prob,
human_probability = 1.0 - new_ai_prob, # Simplified
mixed_probability = result.mixed_probability,
confidence = result.confidence,
details = result.details
)
return calibrated
def _get_domain_performance_weights(self, domain: Domain, metric_names: List[str]) -> Dict[str, float]:
"""
Get domain-specific performance weights (would come from validation data)
"""
# Placeholder - in practice, this would be based on historical performance per domain : FUTURE WORK
performance_weights = {'structural' : 1.0,
'entropy' : 1.0,
'semantic_analysis' : 1.0,
'linguistic' : 1.0,
'perplexity' : 1.0,
'multi_perturbation_stability' : 1.0,
}
# Domain-specific adjustments for all 16 domains
domain_adjustments = {Domain.GENERAL : {'structural' : 1.0,
'perplexity' : 1.0,
'entropy' : 1.0,
'semantic_analysis' : 1.0,
'linguistic' : 1.0,
'multi_perturbation_stability' : 1.0,
},
Domain.ACADEMIC : {'structural' : 1.2,
'perplexity' : 1.3,
'entropy' : 0.9,
'semantic_analysis' : 1.1,
'linguistic' : 1.3,
'multi_perturbation_stability' : 0.8,
},
Domain.CREATIVE : {'structural' : 0.9,
'perplexity' : 1.1,
'entropy' : 1.2,
'semantic_analysis' : 1.0,
'linguistic' : 1.1,
'multi_perturbation_stability' : 0.9,
},
Domain.AI_ML : {'structural' : 1.2,
'perplexity' : 1.3,
'entropy' : 0.9,
'semantic_analysis' : 1.1,
'linguistic' : 1.2,
'multi_perturbation_stability' : 0.8,
},
Domain.SOFTWARE_DEV : {'structural' : 1.2,
'perplexity' : 1.3,
'entropy' : 0.9,
'semantic_analysis' : 1.1,
'linguistic' : 1.2,
'multi_perturbation_stability' : 0.8,
},
Domain.TECHNICAL_DOC : {'structural' : 1.3,
'perplexity' : 1.3,
'entropy' : 0.9,
'semantic_analysis' : 1.2,
'linguistic' : 1.2,
'multi_perturbation_stability' : 0.8,
},
Domain.ENGINEERING : {'structural' : 1.2,
'perplexity' : 1.3,
'entropy' : 0.9,
'semantic_analysis' : 1.1,
'linguistic' : 1.2,
'multi_perturbation_stability' : 0.8,
},
Domain.SCIENCE : {'structural' : 1.2,
'perplexity' : 1.3,
'entropy' : 0.9,
'semantic_analysis' : 1.1,
'linguistic' : 1.2,
'multi_perturbation_stability' : 0.8,
},
Domain.BUSINESS : {'structural' : 1.1,
'perplexity' : 1.2,
'entropy' : 1.0,
'semantic_analysis' : 1.1,
'linguistic' : 1.1,
'multi_perturbation_stability' : 0.9,
},
Domain.LEGAL : {'structural' : 1.3,
'perplexity' : 1.3,
'entropy' : 0.9,
'semantic_analysis' : 1.2,
'linguistic' : 1.3,
'multi_perturbation_stability' : 0.8,
},
Domain.MEDICAL : {'structural' : 1.2,
'perplexity' : 1.3,
'entropy' : 0.9,
'semantic_analysis' : 1.2,
'linguistic' : 1.2,
'multi_perturbation_stability' : 0.8,
},
Domain.JOURNALISM : {'structural' : 1.1,
'perplexity' : 1.2,
'entropy' : 1.0,
'semantic_analysis' : 1.1,
'linguistic' : 1.1,
'multi_perturbation_stability' : 0.8,
},
Domain.MARKETING : {'structural' : 1.0,
'perplexity' : 1.1,
'entropy' : 1.1,
'semantic_analysis' : 1.0,
'linguistic' : 1.2,
'multi_perturbation_stability' : 0.8,
},
Domain.SOCIAL_MEDIA : {'structural' : 0.8,
'perplexity' : 1.0,
'entropy' : 1.3,
'semantic_analysis' : 0.9,
'linguistic' : 0.7,
'multi_perturbation_stability' : 0.9,
},
Domain.BLOG_PERSONAL : {'structural' : 0.9,
'perplexity' : 1.1,
'entropy' : 1.2,
'semantic_analysis' : 1.0,
'linguistic' : 1.0,
'multi_perturbation_stability' : 0.8,
},
Domain.TUTORIAL : {'structural' : 1.1,
'perplexity' : 1.2,
'entropy' : 1.0,
'semantic_analysis' : 1.1,
'linguistic' : 1.1,
'multi_perturbation_stability' : 0.8,
},
}
adjustments = domain_adjustments.get(domain, {})
return {name: performance_weights.get(name, 1.0) * adjustments.get(name, 1.0) for name in metric_names}
def _calculate_consensus_weights(self, results: Dict[str, MetricResult], base_weights: Dict[str, float]) -> Dict[str, float]:
"""
Calculate weights based on metric consensus
"""
# Calculate average AI probability
avg_ai_prob = np.mean([r.ai_probability for r in results.values()])
consensus_weights = dict()
for name, result in results.items():
base_weight = base_weights.get(name, 0.0)
# Reward metrics that agree with consensus
agreement = 1.0 - abs(result.ai_probability - avg_ai_prob)
consensus_weights[name] = base_weight * (0.5 + 0.5 * agreement) # 0.5-1.0 range
# Normalize
total_weight = sum(consensus_weights.values())
if (total_weight > 0):
consensus_weights = {name: w / total_weight for name, w in consensus_weights.items()}
return consensus_weights
def _extract_ml_features(self, results: Dict[str, MetricResult]) -> List[float]:
"""
Extract features for ML ensemble
"""
features = list()
for name in sorted(results.keys()): # Ensure consistent order
result = results[name]
features.extend([result.ai_probability,
result.human_probability,
result.mixed_probability,
result.confidence
])
return features
def _calculate_advanced_confidence(self, results: Dict[str, MetricResult], weights: Dict[str, float], aggregated: Dict[str, float]) -> float:
"""
Calculate advanced confidence considering multiple factors
"""
# Base confidence from metric confidences
base_confidence = sum(result.confidence * weights.get(name, 0.0) for name, result in results.items())
# Agreement factor
ai_probs = [r.ai_probability for r in results.values()]
agreement = 1.0 - min(1.0, np.std(ai_probs) * 2.0) # 0-1 scale
# Certainty factor (how far from 0.5)
certainty = 1.0 - 2.0 * abs(aggregated["ai_probability"] - 0.5)
# Metric quality factor
high_confidence_metrics = sum(1 for r in results.values() if r.confidence > 0.7)
quality_factor = high_confidence_metrics / len(results) if results else 0.0
# Combined confidence
confidence = (base_confidence * 0.4 + agreement * 0.3 + certainty * 0.2 + quality_factor * 0.1)
return max(0.0, min(1.0, confidence))
def _calculate_uncertainty(self, results: Dict[str, MetricResult], weights: Dict[str, float], aggregated: Dict[str, float]) -> float:
"""
Calculate uncertainty score
"""
# Variance in predictions
ai_probs = [r.ai_probability for r in results.values()]
variance_uncertainty = np.var(ai_probs) if len(ai_probs) > 1 else 0.0
# Confidence uncertainty
avg_confidence = np.mean([r.confidence for r in results.values()])
confidence_uncertainty = 1.0 - avg_confidence
# Decision uncertainty (how close to 0.5)
decision_uncertainty = 1.0 - 2.0 * abs(aggregated["ai_probability"] - 0.5)
# Combined uncertainty
uncertainty = (variance_uncertainty * 0.4 + confidence_uncertainty * 0.3 + decision_uncertainty * 0.3)
return max(0.0, min(1.0, uncertainty))
def _calculate_consensus_level(self, results: Dict[str, MetricResult]) -> float:
"""
Calculate consensus level among metrics
"""
if (len(results) < 2):
# Perfect consensus with only one metric
return 1.0
ai_probs = [r.ai_probability for r in results.values()]
std_dev = np.std(ai_probs)
# Convert to consensus level (1.0 = perfect consensus, 0.0 = no consensus)
consensus = 1.0 - min(1.0, std_dev * 2.0)
return consensus
def _apply_adaptive_threshold(self, aggregated: Dict[str, float], base_threshold: float, uncertainty: float) -> str:
"""
Apply adaptive threshold considering uncertainty
"""
ai_prob = aggregated.get("ai_probability", 0.5)
mixed_prob = aggregated.get("mixed_probability", 0.0)
# Adjust threshold based on uncertainty : Higher uncertainty requires more confidence
adjusted_threshold = base_threshold + (uncertainty * 0.1)
# Check for mixed content
if ((mixed_prob > 0.25) or ((uncertainty > 0.6) and (0.3 < ai_prob < 0.7))):
return "Mixed (AI + Human)"
# Apply adjusted threshold
if (ai_prob >= adjusted_threshold):
return "AI-Generated"
elif (ai_prob <= (1.0 - adjusted_threshold)):
return "Human-Written"
else:
return "Uncertain"
def _generate_detailed_reasoning(self, results: Dict[str, MetricResult], weights: Dict[str, float], aggregated: Dict[str, float],
verdict: str, uncertainty: float, consensus: float) -> List[str]:
"""
Generate detailed reasoning for the prediction
"""
reasoning = list()
# Overall assessment
ai_prob = aggregated.get("ai_probability", 0.5)
mixed_prob = aggregated.get("mixed_probability", 0.0)
reasoning.append(f"## Ensemble Analysis Result")
reasoning.append(f"**Final Verdict**: {verdict}")
reasoning.append(f"**AI Probability**: {ai_prob:.1%}")
reasoning.append(f"**Confidence Level**: {self._get_confidence_label(ai_prob)}")
reasoning.append(f"**Uncertainty**: {uncertainty:.1%}")
reasoning.append(f"**Consensus**: {consensus:.1%}")
# Metric analysis
reasoning.append(f"\n## Metric Analysis")
sorted_metrics = sorted(results.items(), key=lambda x: weights.get(x[0], 0.0), reverse=True)
for name, result in sorted_metrics:
weight = weights.get(name, 0.0)
contribution = "High" if (weight > 0.15) else "Medium" if (weight > 0.08) else "Low"
reasoning.append(f"**{name}**: {result.ai_probability:.1%} AI "
f"(Confidence: {result.confidence:.1%}, "
f"Contribution: {contribution})")
# Key factors
reasoning.append(f"\n## Key Decision Factors")
if (uncertainty > 0.7):
reasoning.append("⚠ **High uncertainty** - Metrics show significant disagreement")
elif (consensus > 0.8):
reasoning.append("✓ **Strong consensus** - All metrics agree on classification")
top_metric = sorted_metrics[0] if sorted_metrics else None
if (top_metric and (weights.get(top_metric[0], 0.0) > 0.2)):
reasoning.append(f"🎯 **Dominant metric** - {top_metric[0]} had strongest influence")
if (mixed_prob > 0.2):
reasoning.append("🔀 **Mixed signals** - Content shows characteristics of both AI and human writing")
return reasoning
def _get_confidence_label(self, ai_prob: float) -> str:
"""
Get human-readable confidence label
"""
if ((ai_prob > 0.9) or (ai_prob < 0.1)):
return "Very High"
elif ((ai_prob > 0.8) or (ai_prob < 0.2)):
return "High"
elif ((ai_prob > 0.7) or (ai_prob < 0.3)):
return "Moderate"
else:
return "Low"
def _create_fallback_result(self, domain: Domain, metric_results: Dict[str, MetricResult], error: str) -> EnsembleResult:
"""
Create fallback result when ensemble cannot make a confident decision
"""
return EnsembleResult(final_verdict = "Uncertain",
ai_probability = 0.5,
human_probability = 0.5,
mixed_probability = 0.0,
overall_confidence = 0.0,
domain = domain,
metric_results = metric_results,
metric_weights = {},
weighted_scores = {},
reasoning = [f"Ensemble analysis inconclusive", f"Reason: {error}"],
uncertainty_score = 1.0,
consensus_level = 0.0,
)
# Export
__all__ = ["EnsembleResult",
"EnsembleClassifier",
] |