File size: 17,908 Bytes
8f8d0f6
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
import re
import string
import logging
from typing import Dict, List, Any, Optional
import pandas as pd
import numpy as np
from collections import Counter

# NLTK imports
import nltk
try:
    from nltk.sentiment import SentimentIntensityAnalyzer
    from nltk.corpus import stopwords
    from nltk.tokenize import word_tokenize, sent_tokenize
    from nltk.stem import PorterStemmer
except ImportError:
    pass

# Download required NLTK data
try:
    nltk.download('vader_lexicon', quiet=True)
    nltk.download('punkt', quiet=True)
    nltk.download('stopwords', quiet=True)
except:
    pass

# Transformers for FinBERT
try:
    from transformers import AutoTokenizer, AutoModelForSequenceClassification, pipeline
    import torch
except ImportError:
    pass

# YAKE for keyword extraction
try:
    import yake
except ImportError:
    pass

logger = logging.getLogger(__name__)

class SentimentAnalyzer:
    """Multi-model sentiment analysis"""
    
    def __init__(self):
        self.vader_analyzer = None
        self.finbert_pipeline = None
        self.loughran_mcdonald_dict = None
        
        self._initialize_models()
        logger.info("SentimentAnalyzer initialized")
    
    def _initialize_models(self):
        """Initialize all sentiment analysis models"""
        # VADER
        try:
            self.vader_analyzer = SentimentIntensityAnalyzer()
            logger.info("VADER model loaded")
        except Exception as e:
            logger.error(f"Failed to load VADER: {str(e)}")
        
        # FinBERT
        try:
            model_name = "ProsusAI/finbert"
            self.finbert_pipeline = pipeline(
                "sentiment-analysis",
                model=model_name,
                tokenizer=model_name,
                device=0 if torch.cuda.is_available() else -1
            )
            logger.info("FinBERT model loaded")
        except Exception as e:
            logger.warning(f"Failed to load FinBERT, using CPU fallback: {str(e)}")
            try:
                model_name = "ProsusAI/finbert"
                self.finbert_pipeline = pipeline(
                    "sentiment-analysis",
                    model=model_name,
                    tokenizer=model_name,
                    device=-1
                )
                logger.info("FinBERT model loaded on CPU")
            except Exception as e2:
                logger.error(f"Failed to load FinBERT completely: {str(e2)}")
        
        # Loughran-McDonald Dictionary
        try:
            self.loughran_mcdonald_dict = self._load_loughran_mcdonald()
            logger.info("Loughran-McDonald dictionary loaded")
        except Exception as e:
            logger.error(f"Failed to load Loughran-McDonald dictionary: {str(e)}")
    
    def _load_loughran_mcdonald(self) -> Dict[str, List[str]]:
        """Load Loughran-McDonald financial sentiment dictionary"""
        # Simplified version with key financial sentiment words
        return {
            'positive': [
                'profit', 'profitable', 'profitability', 'revenue', 'revenues', 'growth', 
                'growing', 'increase', 'increased', 'increasing', 'success', 'successful',
                'gain', 'gains', 'benefit', 'benefits', 'improvement', 'improved', 'strong',
                'stronger', 'excellent', 'outstanding', 'exceed', 'exceeded', 'exceeds',
                'beat', 'beats', 'positive', 'optimistic', 'bullish', 'rise', 'rising',
                'surge', 'surged', 'boom', 'booming', 'expand', 'expansion', 'opportunity',
                'opportunities', 'advance', 'advances', 'achievement', 'achieve', 'winner'
            ],
            'negative': [
                'loss', 'losses', 'lose', 'losing', 'decline', 'declining', 'decrease',
                'decreased', 'decreasing', 'fall', 'falling', 'drop', 'dropped', 'plunge',
                'plunged', 'crash', 'crashed', 'failure', 'failed', 'weak', 'weakness',
                'poor', 'worse', 'worst', 'bad', 'terrible', 'crisis', 'problem', 'problems',
                'risk', 'risks', 'risky', 'concern', 'concerns', 'worried', 'worry',
                'negative', 'pessimistic', 'bearish', 'bankruptcy', 'bankrupt', 'deficit',
                'debt', 'lawsuit', 'sue', 'sued', 'investigation', 'fraud', 'scandal',
                'volatility', 'volatile', 'uncertainty', 'uncertain', 'challenge', 'challenges'
            ]
        }
    
    def analyze_sentiment(self, text: str, models: List[str] = None) -> Dict[str, Any]:
        """Analyze sentiment using multiple models"""
        if models is None:
            models = ['VADER', 'Loughran-McDonald', 'FinBERT']
        
        results = {}
        
        # Clean text
        cleaned_text = self._clean_text(text)
        
        # VADER Analysis
        if 'VADER' in models and self.vader_analyzer:
            try:
                vader_scores = self.vader_analyzer.polarity_scores(cleaned_text)
                results['vader'] = vader_scores['compound']
                results['vader_detailed'] = vader_scores
            except Exception as e:
                logger.error(f"VADER analysis failed: {str(e)}")
                results['vader'] = 0.0
        
        # Loughran-McDonald Analysis
        if 'Loughran-McDonald' in models and self.loughran_mcdonald_dict:
            try:
                lm_score = self._analyze_loughran_mcdonald(cleaned_text)
                results['loughran_mcdonald'] = lm_score
            except Exception as e:
                logger.error(f"Loughran-McDonald analysis failed: {str(e)}")
                results['loughran_mcdonald'] = 0.0
        
        # FinBERT Analysis
        if 'FinBERT' in models and self.finbert_pipeline:
            try:
                # Truncate text for FinBERT (max 512 tokens)
                truncated_text = cleaned_text[:2000]  # Approximate token limit
                finbert_result = self.finbert_pipeline(truncated_text)[0]
                
                # Convert to numerical score
                label = finbert_result['label'].lower()
                confidence = finbert_result['score']
                
                if label == 'positive':
                    finbert_score = confidence
                elif label == 'negative':
                    finbert_score = -confidence
                else:  # neutral
                    finbert_score = 0.0
                
                results['finbert'] = finbert_score
                results['finbert_detailed'] = finbert_result
                
            except Exception as e:
                logger.error(f"FinBERT analysis failed: {str(e)}")
                results['finbert'] = 0.0
        
        # Calculate composite score
        scores = []
        weights = {'vader': 0.3, 'loughran_mcdonald': 0.4, 'finbert': 0.3}
        
        for model in ['vader', 'loughran_mcdonald', 'finbert']:
            if model in results:
                scores.append(results[model] * weights[model])
        
        results['compound'] = sum(scores) if scores else 0.0
        
        return results
    
    def _clean_text(self, text: str) -> str:
        """Clean text for sentiment analysis"""
        if not text:
            return ""
        
        # Remove URLs
        text = re.sub(r'http[s]?://(?:[a-zA-Z]|[0-9]|[$-_@.&+]|[!*\\(\\),]|(?:%[0-9a-fA-F][0-9a-fA-F]))+', '', text)
        
        # Remove email addresses
        text = re.sub(r'\b[A-Za-z0-9._%+-]+@[A-Za-z0-9.-]+\.[A-Z|a-z]{2,}\b', '', text)
        
        # Remove extra whitespace
        text = re.sub(r'\s+', ' ', text)
        
        # Remove special characters but keep basic punctuation
        text = re.sub(r'[^\w\s.,!?;:\-\'"()]', '', text)
        
        return text.strip()
    
    def _analyze_loughran_mcdonald(self, text: str) -> float:
        """Analyze sentiment using Loughran-McDonald dictionary"""
        try:
            words = word_tokenize(text.lower())
            
            positive_count = sum(1 for word in words if word in self.loughran_mcdonald_dict['positive'])
            negative_count = sum(1 for word in words if word in self.loughran_mcdonald_dict['negative'])
            
            total_sentiment_words = positive_count + negative_count
            
            if total_sentiment_words == 0:
                return 0.0
            
            # Calculate normalized score
            score = (positive_count - negative_count) / len(words) * 10  # Scale factor
            
            # Clamp to [-1, 1] range
            return max(-1.0, min(1.0, score))
            
        except Exception as e:
            logger.error(f"Loughran-McDonald calculation error: {str(e)}")
            return 0.0

class KeywordExtractor:
    """Extract important keywords from text using YAKE"""
    
    def __init__(self):
        self.stop_words = set()
        try:
            self.stop_words = set(stopwords.words('english'))
        except:
            # Fallback stop words
            self.stop_words = {
                'the', 'a', 'an', 'and', 'or', 'but', 'in', 'on', 'at', 'to', 'for', 
                'of', 'with', 'by', 'is', 'are', 'was', 'were', 'be', 'been', 'have',
                'has', 'had', 'do', 'does', 'did', 'will', 'would', 'could', 'should',
                'may', 'might', 'must', 'can', 'this', 'that', 'these', 'those'
            }
        
        logger.info("KeywordExtractor initialized")
    
    def extract_keywords(self, text: str, num_keywords: int = 20) -> List[Dict[str, Any]]:
        """Extract keywords using YAKE algorithm"""
        try:
            # Use YAKE if available
            if 'yake' in globals():
                return self._extract_with_yake(text, num_keywords)
            else:
                return self._extract_with_frequency(text, num_keywords)
                
        except Exception as e:
            logger.error(f"Keyword extraction failed: {str(e)}")
            return []
    
    def _extract_with_yake(self, text: str, num_keywords: int) -> List[Dict[str, Any]]:
        """Extract keywords using YAKE algorithm"""
        try:
            # YAKE configuration
            kw_extractor = yake.KeywordExtractor(
                lan="en",
                n=3,  # n-gram size
                dedupLim=0.9,
                top=num_keywords,
                features=None
            )
            
            keywords = kw_extractor.extract_keywords(text)
            
            # Convert to desired format (lower score = more relevant in YAKE)
            result = []
            for keyword, score in keywords:
                result.append({
                    'keyword': keyword,
                    'score': 1.0 / (1.0 + score),  # Invert score so higher = more relevant
                    'relevance': 'high' if score < 0.1 else 'medium' if score < 0.3 else 'low'
                })
            
            return result
            
        except Exception as e:
            logger.error(f"YAKE extraction failed: {str(e)}")
            return self._extract_with_frequency(text, num_keywords)
    
    def _extract_with_frequency(self, text: str, num_keywords: int) -> List[Dict[str, Any]]:
        """Fallback keyword extraction using frequency analysis"""
        try:
            # Clean and tokenize
            words = word_tokenize(text.lower())
            
            # Filter words
            filtered_words = [
                word for word in words 
                if (word not in self.stop_words and 
                    word not in string.punctuation and 
                    len(word) > 2 and 
                    word.isalpha())
            ]
            
            # Count frequencies
            word_freq = Counter(filtered_words)
            
            # Get top keywords
            top_words = word_freq.most_common(num_keywords)
            
            # Calculate relevance scores
            max_freq = top_words[0][1] if top_words else 1
            
            result = []
            for word, freq in top_words:
                score = freq / max_freq
                result.append({
                    'keyword': word,
                    'score': score,
                    'relevance': 'high' if score > 0.7 else 'medium' if score > 0.3 else 'low'
                })
            
            return result
            
        except Exception as e:
            logger.error(f"Frequency extraction failed: {str(e)}")
            return []

class TextProcessor:
    """Text preprocessing and cleaning utilities"""
    
    def __init__(self):
        self.stemmer = PorterStemmer()
        logger.info("TextProcessor initialized")
    
    def clean_article_content(self, content: str) -> str:
        """Clean article content by removing boilerplate"""
        if not content:
            return ""
        
        # Remove common boilerplate patterns
        boilerplate_patterns = [
            r'Subscribe to our newsletter.*',
            r'Sign up for.*',
            r'Follow us on.*',
            r'Copyright.*',
            r'All rights reserved.*',
            r'Terms of use.*',
            r'Privacy policy.*',
            r'Cookie policy.*',
            r'\d+ comments?',
            r'Share this article.*',
            r'Related articles?.*',
            r'More from.*',
            r'Advertisement.*',
            r'Sponsored content.*'
        ]
        
        cleaned_content = content
        for pattern in boilerplate_patterns:
            cleaned_content = re.sub(pattern, '', cleaned_content, flags=re.IGNORECASE)
        
        # Remove extra whitespace
        cleaned_content = re.sub(r'\s+', ' ', cleaned_content)
        
        # Remove very short sentences (likely navigation/boilerplate)
        sentences = sent_tokenize(cleaned_content)
        meaningful_sentences = [
            sent for sent in sentences 
            if len(sent.split()) > 5 and not self._is_boilerplate_sentence(sent)
        ]
        
        return ' '.join(meaningful_sentences).strip()
    
    def _is_boilerplate_sentence(self, sentence: str) -> bool:
        """Check if sentence is likely boilerplate"""
        boilerplate_indicators = [
            'click here', 'read more', 'subscribe', 'follow us', 'contact us',
            'terms of service', 'privacy policy', 'copyright', 'all rights reserved',
            'advertisement', 'sponsored', 'related articles'
        ]
        
        sentence_lower = sentence.lower()
        return any(indicator in sentence_lower for indicator in boilerplate_indicators)
    
    def extract_entities(self, text: str) -> Dict[str, List[str]]:
        """Extract named entities (companies, people, locations)"""
        # Simple regex-based entity extraction
        entities = {
            'companies': [],
            'people': [],
            'locations': [],
            'money': [],
            'dates': []
        }
        
        try:
            # Company patterns (simplified)
            company_pattern = r'\b[A-Z][a-zA-Z]+ (?:Inc|Corp|LLC|Ltd|Company|Co)\b'
            entities['companies'] = list(set(re.findall(company_pattern, text)))
            
            # Money patterns
            money_pattern = r'\$[\d,]+(?:\.\d{2})?(?:\s?(?:million|billion|trillion|k|M|B|T))?'
            entities['money'] = list(set(re.findall(money_pattern, text)))
            
            # Date patterns (simplified)
            date_pattern = r'\b(?:January|February|March|April|May|June|July|August|September|October|November|December)\s+\d{1,2},?\s+\d{4}'
            entities['dates'] = list(set(re.findall(date_pattern, text)))
            
        except Exception as e:
            logger.error(f"Entity extraction failed: {str(e)}")
        
        return entities
    
    def calculate_readability(self, text: str) -> Dict[str, float]:
        """Calculate text readability metrics"""
        try:
            sentences = sent_tokenize(text)
            words = word_tokenize(text)
            
            if not sentences or not words:
                return {'flesch_score': 0.0, 'avg_sentence_length': 0.0, 'avg_word_length': 0.0}
            
            # Basic metrics
            num_sentences = len(sentences)
            num_words = len(words)
            num_syllables = sum(self._count_syllables(word) for word in words if word.isalpha())
            
            # Average sentence length
            avg_sentence_length = num_words / num_sentences
            
            # Average word length
            avg_word_length = sum(len(word) for word in words if word.isalpha()) / num_words
            
            # Flesch Reading Ease Score (simplified)
            flesch_score = 206.835 - (1.015 * avg_sentence_length) - (84.6 * (num_syllables / num_words))
            
            return {
                'flesch_score': max(0.0, min(100.0, flesch_score)),
                'avg_sentence_length': avg_sentence_length,
                'avg_word_length': avg_word_length
            }
            
        except Exception as e:
            logger.error(f"Readability calculation failed: {str(e)}")
            return {'flesch_score': 0.0, 'avg_sentence_length': 0.0, 'avg_word_length': 0.0}
    
    def _count_syllables(self, word: str) -> int:
        """Count syllables in a word (simplified)"""
        word = word.lower()
        vowels = 'aeiouy'
        syllable_count = 0
        prev_char_was_vowel = False
        
        for char in word:
            if char in vowels:
                if not prev_char_was_vowel:
                    syllable_count += 1
                prev_char_was_vowel = True
            else:
                prev_char_was_vowel = False
        
        # Handle silent e
        if word.endswith('e'):
            syllable_count -= 1
        
        # Every word has at least one syllable
        return max(1, syllable_count)