Update app.py

app.py

@@ -7,6 +7,14 @@ from plotly.subplots import make_subplots
 import numpy as np
 from wordcloud import WordCloud
 from collections import Counter, defaultdict
+import re
+import json
+import csv
+import io
+import tempfile
+from datetime import datetime
+import logging
+from functools import lru_cache, wraps
 from dataclasses import dataclass
 from typing import List, Dict, Optional, Tuple, Any, Callable
 from contextlib import contextmanager
@@ -16,7 +24,18 @@ import langdetect
 import pandas as pd
 import gc
 
+# Advanced analysis imports
+import shap
+import lime
+from lime.lime_text import LimeTextExplainer
+
 # Configuration
+@dataclass
+class Config:
+    MAX_HISTORY_SIZE: int = 1000
+    BATCH_SIZE_LIMIT: int = 50
+    MAX_TEXT_LENGTH: int = 512
+    MIN_WORD_LENGTH: int = 2
     CACHE_SIZE: int = 128
     BATCH_PROCESSING_SIZE: int = 8
     
@@ -35,7 +54,6 @@ import gc
         'en': "cardiffnlp/twitter-roberta-base-sentiment-latest",
         'multilingual': "cardiffnlp/twitter-xlm-roberta-base-sentiment",
         'zh': "uer/roberta-base-finetuned-dianping-chinese"
-
     }
     
     # Color themes for Plotly
@@ -63,19 +81,28 @@ except:
 # Decorators and Context Managers
 def handle_errors(default_return=None):
     """Centralized error handling decorator"""
+    def decorator(func: Callable) -> Callable:
+        @wraps(func)
+        def wrapper(*args, **kwargs):
+            try:
+                return func(*args, **kwargs)
+            except Exception as e:
+                logger.error(f"{func.__name__} failed: {e}")
+                return default_return if default_return is not None else f"Error: {str(e)}"
+        return wrapper
     return decorator
 
 @contextmanager
 def memory_cleanup():
     """Context manager for memory cleanup"""
-
     try:
         yield
     finally:
-
         gc.collect()
 
 class ThemeContext:
+    """Theme management context"""
+    def __init__(self, theme: str = 'default'):
         self.theme = theme
         self.colors = config.THEMES.get(theme, config.THEMES['default'])
 
@@ -83,9 +110,6 @@ class ThemeContext:
 class ModelManager:
     """Multi-language model manager with lazy loading"""
     _instance = None
-
-
-
     
     def __new__(cls):
         if cls._instance is None:
@@ -103,16 +127,6 @@ class ModelManager:
     
     def _load_default_models(self):
         """Load default models"""
-
-
-
-
-
-
-
-
-
-
         try:
             # Load multilingual model as default
             model_name = config.MODELS['multilingual']
@@ -241,6 +255,7 @@ class HistoryManager:
     def clear(self) -> int:
         count = len(self._history)
         self._history.clear()
+        return count
     
     def size(self) -> int:
         return len(self._history)
@@ -266,93 +281,16 @@ class HistoryManager:
             'most_common_language': Counter(languages).most_common(1)[0][0] if languages else 'en'
         }
 
-# Core Sentiment Analysis Engine
+# Core Sentiment Analysis Engine (Modified - removed attention analysis)
 class SentimentEngine:
     """Multi-language sentiment analysis engine"""
     
     def __init__(self):
         self.model_manager = ModelManager()
     
-    def extract_attention_keywords(self, text: str, language: str = 'auto', top_k: int = 10) -> List[Tuple[str, float]]:
-        """Extract keywords using attention weights"""
-        try:
-            if language == 'auto':
-                language = self.model_manager.detect_language(text)
-            
-            model, tokenizer = self.model_manager.get_model(language)
-            
-            inputs = tokenizer(
-                text, return_tensors="pt", padding=True, 
-                truncation=True, max_length=config.MAX_TEXT_LENGTH
-            ).to(self.model_manager.device)
-            
-
-            with torch.no_grad():
-                outputs = model(**inputs, output_attentions=True)
-
-                
-                if hasattr(outputs, 'attentions') and outputs.attentions:
-                    # Use attention weights
-                    attention = outputs.attentions[-1]
-                    avg_attention = attention.mean(dim=1)[0, 0, :]
-                    
-                    tokens = tokenizer.convert_ids_to_tokens(inputs['input_ids'][0])
-                    attention_scores = avg_attention.cpu().numpy()
-                    
-                    # Process tokens and scores
-                    word_scores = {}
-                    current_word = ""
-                    current_score = 0.0
-                    
-                    for token, score in zip(tokens, attention_scores):
-                        if token in ['[CLS]', '[SEP]', '[PAD]', '<s>', '</s>']:
-                            continue
-                            
-                        if token.startswith('##') or token.startswith('▁'):
-                            current_word += token.replace('##', '').replace('▁', '')
-                            current_score = max(current_score, score)
-                        else:
-                            if current_word and len(current_word) >= config.MIN_WORD_LENGTH:
-                                word_scores[current_word.lower()] = current_score
-                            current_word = token
-                            current_score = score
-                    
-
-
-
-
-
-
-                    if current_word and len(current_word) >= config.MIN_WORD_LENGTH:
-                        word_scores[current_word.lower()] = current_score
-                    
-                    # Filter and sort
-                    filtered_words = {
-                        word: score for word, score in word_scores.items() 
-                        if word not in STOP_WORDS and len(word) >= config.MIN_WORD_LENGTH
-                    }
-                    
-                    sorted_words = sorted(filtered_words.items(), key=lambda x: x[1], reverse=True)
-                    return sorted_words[:top_k]
-                
-
-
-
-
-
-
-
-
-        except Exception as e:
-            logger.error(f"Attention keyword extraction failed: {e}")
-        
-        # Fallback to simple keyword extraction
-        keywords = TextProcessor.extract_keywords(text, top_k)
-        return [(word, 0.1) for word in keywords]
-    
     @handle_errors(default_return={'sentiment': 'Unknown', 'confidence': 0.0, 'keywords': []})
     def analyze_single(self, text: str, language: str = 'auto', preprocessing_options: Dict = None) -> Dict:
-        """Analyze single text with enhanced features"""
+        """Analyze single text with basic features"""
         if not text.strip():
             raise ValueError("Empty text provided")
         
@@ -378,7 +316,6 @@ class SentimentEngine:
         # Tokenize and analyze
         inputs = tokenizer(processed_text, return_tensors="pt", padding=True, 
                          truncation=True, max_length=config.MAX_TEXT_LENGTH).to(self.model_manager.device)
-
         
         with torch.no_grad():
             outputs = model(**inputs)
@@ -413,13 +350,14 @@ class SentimentEngine:
                 'has_neutral': False
             }
         
-        # Extract keywords
-        keywords = self.extract_attention_keywords(text, detected_lang)
+        # Extract basic keywords
+        keywords = TextProcessor.extract_keywords(text, 10)
+        keyword_tuples = [(word, 0.1) for word in keywords]  # Simple keyword extraction
         
         # Add metadata
         result.update({
             'language': detected_lang,
-            'keywords': keywords,
+            'keywords': keyword_tuples,
             'word_count': len(text.split()),
             'char_count': len(text)
         })
@@ -433,6 +371,12 @@ class SentimentEngine:
         if len(texts) > config.BATCH_SIZE_LIMIT:
             texts = texts[:config.BATCH_SIZE_LIMIT]
         
+        results = []
+        batch_size = config.BATCH_PROCESSING_SIZE
+        
+        for i in range(0, len(texts), batch_size):
+            batch = texts[i:i+batch_size]
+            
             if progress_callback:
                 progress_callback((i + len(batch)) / len(texts))
             
@@ -452,17 +396,191 @@ class SentimentEngine:
                         'text': text[:100] + '...' if len(text) > 100 else text,
                         'full_text': text
                     })
-
-
-
-
-
-
-
         
         return results
 
-# Advanced Plotly Visualization System
+# Advanced Analysis Engine (NEW)
+class AdvancedAnalysisEngine:
+    """Advanced analysis using SHAP and LIME"""
+    
+    def __init__(self):
+        self.model_manager = ModelManager()
+    
+    def create_prediction_function(self, model, tokenizer, device):
+        """Create prediction function for LIME/SHAP"""
+        def predict_proba(texts):
+            results = []
+            for text in texts:
+                inputs = tokenizer(text, return_tensors="pt", padding=True, 
+                                 truncation=True, max_length=config.MAX_TEXT_LENGTH).to(device)
+                with torch.no_grad():
+                    outputs = model(**inputs)
+                    probs = torch.nn.functional.softmax(outputs.logits, dim=-1).cpu().numpy()[0]
+                results.append(probs)
+            return np.array(results)
+        return predict_proba
+    
+    @handle_errors(default_return=("Analysis failed", None, None))
+    def analyze_with_shap(self, text: str, language: str = 'auto') -> Tuple[str, go.Figure, Dict]:
+        """Perform SHAP analysis"""
+        if not text.strip():
+            return "Please enter text for analysis", None, {}
+        
+        # Detect language and get model
+        if language == 'auto':
+            detected_lang = self.model_manager.detect_language(text)
+        else:
+            detected_lang = language
+        
+        model, tokenizer = self.model_manager.get_model(detected_lang)
+        
+        # Create prediction function
+        predict_fn = self.create_prediction_function(model, tokenizer, self.model_manager.device)
+        
+        try:
+            # Initialize SHAP explainer
+            explainer = shap.Explainer(predict_fn, tokenizer)
+            
+            # Get SHAP values
+            shap_values = explainer([text])
+            
+            # Extract token importance
+            tokens = shap_values.data[0]
+            values = shap_values.values[0]
+            
+            # Create visualization data
+            if len(values.shape) > 1:
+                # Multi-class case
+                pos_values = values[:, -1] if values.shape[1] == 3 else values[:, 1]
+            else:
+                pos_values = values
+            
+            # Create SHAP plot
+            fig = go.Figure()
+            
+            colors = ['red' if v < 0 else 'green' for v in pos_values]
+            
+            fig.add_trace(go.Bar(
+                x=list(range(len(tokens))),
+                y=pos_values,
+                text=tokens,
+                textposition='outside',
+                marker_color=colors,
+                name='SHAP Values'
+            ))
+            
+            fig.update_layout(
+                title="SHAP Analysis - Token Importance",
+                xaxis_title="Token Index",
+                yaxis_title="SHAP Value",
+                height=500,
+                xaxis=dict(tickmode='array', tickvals=list(range(len(tokens))), ticktext=tokens)
+            )
+            
+            # Create analysis summary
+            analysis_data = {
+                'method': 'SHAP',
+                'language': detected_lang,
+                'total_tokens': len(tokens),
+                'positive_influence': sum(1 for v in pos_values if v > 0),
+                'negative_influence': sum(1 for v in pos_values if v < 0),
+                'most_important_tokens': [(tokens[i], float(pos_values[i])) 
+                                        for i in np.argsort(np.abs(pos_values))[-5:]]
+            }
+            
+            summary_text = f"""
+**SHAP Analysis Results:**
+- **Language:** {detected_lang.upper()}
+- **Total Tokens:** {analysis_data['total_tokens']}
+- **Positive Influence Tokens:** {analysis_data['positive_influence']}
+- **Negative Influence Tokens:** {analysis_data['negative_influence']}
+- **Most Important Tokens:** {', '.join([f"{token}({score:.3f})" for token, score in analysis_data['most_important_tokens']])}
+            """
+            
+            return summary_text, fig, analysis_data
+            
+        except Exception as e:
+            logger.error(f"SHAP analysis failed: {e}")
+            return f"SHAP analysis failed: {str(e)}", None, {}
+    
+    @handle_errors(default_return=("Analysis failed", None, None))
+    def analyze_with_lime(self, text: str, language: str = 'auto') -> Tuple[str, go.Figure, Dict]:
+        """Perform LIME analysis"""
+        if not text.strip():
+            return "Please enter text for analysis", None, {}
+        
+        # Detect language and get model
+        if language == 'auto':
+            detected_lang = self.model_manager.detect_language(text)
+        else:
+            detected_lang = language
+        
+        model, tokenizer = self.model_manager.get_model(detected_lang)
+        
+        # Create prediction function
+        predict_fn = self.create_prediction_function(model, tokenizer, self.model_manager.device)
+        
+        try:
+            # Initialize LIME explainer
+            explainer = LimeTextExplainer(class_names=['Negative', 'Neutral', 'Positive'])
+            
+            # Get LIME explanation
+            exp = explainer.explain_instance(text, predict_fn, num_features=20)
+            
+            # Extract feature importance
+            lime_data = exp.as_list()
+            
+            # Create visualization
+            words = [item[0] for item in lime_data]
+            scores = [item[1] for item in lime_data]
+            
+            fig = go.Figure()
+            
+            colors = ['red' if s < 0 else 'green' for s in scores]
+            
+            fig.add_trace(go.Bar(
+                y=words,
+                x=scores,
+                orientation='h',
+                marker_color=colors,
+                text=[f'{s:.3f}' for s in scores],
+                textposition='auto',
+                name='LIME Importance'
+            ))
+            
+            fig.update_layout(
+                title="LIME Analysis - Feature Importance",
+                xaxis_title="Importance Score",
+                yaxis_title="Words/Phrases",
+                height=500
+            )
+            
+            # Create analysis summary
+            analysis_data = {
+                'method': 'LIME',
+                'language': detected_lang,
+                'features_analyzed': len(lime_data),
+                'positive_features': sum(1 for _, score in lime_data if score > 0),
+                'negative_features': sum(1 for _, score in lime_data if score < 0),
+                'feature_importance': lime_data
+            }
+            
+            summary_text = f"""
+**LIME Analysis Results:**
+- **Language:** {detected_lang.upper()}
+- **Features Analyzed:** {analysis_data['features_analyzed']}
+- **Positive Features:** {analysis_data['positive_features']}
+- **Negative Features:** {analysis_data['negative_features']}
+- **Top Features:** {', '.join([f"{word}({score:.3f})" for word, score in lime_data[:5]])}
+            """
+            
+            return summary_text, fig, analysis_data
+            
+        except Exception as e:
+            logger.error(f"LIME analysis failed: {e}")
+            return f"LIME analysis failed: {str(e)}", None, {}
+
+# Advanced Plotly Visualization System (Updated - removed attention visualization)
 class PlotlyVisualizer:
     """Enhanced Plotly visualizations"""
     
@@ -547,31 +665,12 @@ class PlotlyVisualizer:
     @staticmethod
     @handle_errors(default_return=None)
     def create_keyword_chart(keywords: List[Tuple[str, float]], sentiment: str, theme: ThemeContext) -> go.Figure:
-        """Create keyword importance chart"""
+        """Create basic keyword chart"""
         if not keywords:
             fig = go.Figure()
             fig.add_annotation(text="No keywords extracted", 
                              xref="paper", yref="paper", x=0.5, y=0.5, showarrow=False)
             fig.update_layout(height=400, title="Keywords")
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
             return fig
         
         words = [word for word, score in keywords]
@@ -592,7 +691,7 @@ class PlotlyVisualizer:
         
         fig.update_layout(
             title=f"Top Keywords ({sentiment})",
-            xaxis_title="Attention Weight",
+            xaxis_title="Frequency Score",
             yaxis_title="Keywords",
             height=400,
             showlegend=False
@@ -625,14 +724,6 @@ class PlotlyVisualizer:
         )
         
         return fig
-
-
-
-
-
-
-
-
     
     @staticmethod
     @handle_errors(default_return=None)
@@ -730,6 +821,8 @@ class DataHandler:
         if not data:
             return None, "No data to export"
         
+        temp_file = tempfile.NamedTemporaryFile(mode='w', delete=False, 
+                                               suffix=f'.{format_type}', encoding='utf-8')
         
         if format_type == 'csv':
             writer = csv.writer(temp_file)
@@ -751,10 +844,10 @@ class DataHandler:
                 ])
         elif format_type == 'json':
             json.dump(data, temp_file, indent=2, ensure_ascii=False)
+        
         temp_file.close()
         return temp_file.name, f"Exported {len(data)} entries"
     
-
     @staticmethod
     @handle_errors(default_return="")
     def process_file(file) -> str:
@@ -765,7 +858,6 @@ class DataHandler:
         content = file.read().decode('utf-8')
         
         if file.name.endswith('.csv'):
-
             csv_file = io.StringIO(content)
             reader = csv.reader(csv_file)
             try:
@@ -782,6 +874,7 @@ class DataHandler:
                 texts = []
                 for line in lines:
                     if line.strip():
+                        text = line.strip().strip('"')
                         if text:
                             texts.append(text)
                 return '\n'.join(texts)
@@ -794,6 +887,7 @@ class SentimentApp:
     
     def __init__(self):
         self.engine = SentimentEngine()
+        self.advanced_engine = AdvancedAnalysisEngine()  # NEW
         self.history = HistoryManager()
         self.data_handler = DataHandler()
         
@@ -805,12 +899,11 @@ class SentimentApp:
             ["Esta película fue increíble, me encantó la cinematografía."],  # Spanish
             ["Ce film était magnifique, j'ai adoré la réalisation."],  # French
         ]
-
     
     @handle_errors(default_return=("Please enter text", None, None, None))
     def analyze_single(self, text: str, language: str, theme: str, clean_text: bool, 
                        remove_punct: bool, remove_nums: bool):
-        """Single text analysis with enhanced visualizations"""
+        """Single text analysis with basic visualizations (removed attention analysis)"""
         if not text.strip():
             return "Please enter text", None, None, None
         
@@ -966,6 +1059,23 @@ class SentimentApp:
             
             return summary_text, df, summary_fig, confidence_fig
     
+    # NEW: Advanced analysis methods
+    @handle_errors(default_return=("Please enter text", None))
+    def analyze_with_shap(self, text: str, language: str):
+        """Perform SHAP analysis"""
+        language_map = {v: k for k, v in config.SUPPORTED_LANGUAGES.items()}
+        language_code = language_map.get(language, 'auto')
+        
+        return self.advanced_engine.analyze_with_shap(text, language_code)
+    
+    @handle_errors(default_return=("Please enter text", None))
+    def analyze_with_lime(self, text: str, language: str):
+        """Perform LIME analysis"""
+        language_map = {v: k for k, v in config.SUPPORTED_LANGUAGES.items()}
+        language_code = language_map.get(language, 'auto')
+        
+        return self.advanced_engine.analyze_with_lime(text, language_code)
+    
     @handle_errors(default_return=(None, "No history available"))
     def plot_history(self, theme: str = 'default'):
         """Plot comprehensive history analysis"""
@@ -973,9 +1083,7 @@ class SentimentApp:
         if len(history) < 2:
             return None, f"Need at least 2 analyses for trends. Current: {len(history)}"
         
-
         theme_ctx = ThemeContext(theme)
-
         
         with memory_cleanup():
             fig = PlotlyVisualizer.create_history_dashboard(history, theme_ctx)
@@ -1012,9 +1120,9 @@ class SentimentApp:
 - **Languages Detected:** {stats['languages_detected']}
         """
 
-# Gradio Interface
+# Gradio Interface (Updated with Advanced Analysis tab)
 def create_interface():
-    """Create comprehensive Gradio interface"""
+    """Create comprehensive Gradio interface with Advanced Analysis tab"""
     app = SentimentApp()
     
     with gr.Blocks(theme=gr.themes.Soft(), title="Multilingual Sentiment Analyzer") as demo:
@@ -1063,351 +1171,42 @@ def create_interface():
                 probability_plot = gr.Plot(label="Probability Distribution")
             
             with gr.Row():
-                keyword_plot = gr.Plot(label="Key Contributing Words")
+                keyword_plot = gr.Plot(label="Basic Keywords")
         
-        with gr.Tab("Batch Analysis"):
-            with gr.Row():
-                with gr.Column():
-                    file_upload = gr.File(
-                        label="Upload File (CSV/TXT)", 
-                        file_types=[".csv", ".txt"]
-                    )
-                    batch_input = gr.Textbox(
-                        label="Batch Input (one text per line)",
-                        placeholder="Enter multiple texts, one per line...",
-                        lines=10
-                    )
-                    
-                    with gr.Row():
-                        batch_language = gr.Dropdown(
-                            choices=list(config.SUPPORTED_LANGUAGES.values()),
-                            value="Auto Detect",
-                            label="Language"
-                        )
-                        batch_theme = gr.Dropdown(
-                            choices=list(config.THEMES.keys()),
-                            value="default",
-                            label="Theme"
-                        )
-                    
-                    with gr.Row():
-                        batch_clean_cb = gr.Checkbox(label="Clean Text", value=False)
-                        batch_punct_cb = gr.Checkbox(label="Remove Punctuation", value=False)
-                        batch_nums_cb = gr.Checkbox(label="Remove Numbers", value=False)
-                    
-                    with gr.Row():
-                        load_file_btn = gr.Button("Load File")
-                        analyze_batch_btn = gr.Button("Analyze Batch", variant="primary")
-                
-                with gr.Column():
-                    batch_summary = gr.Textbox(label="Batch Summary", lines=8)
-                    batch_results_df = gr.Dataframe(
-                        label="Detailed Results",
-                        headers=["Index", "Text", "Sentiment", "Confidence", "Language", "Keywords"],
-                        datatype=["number", "str", "str", "str", "str", "str"]
-                    )
+        # NEW: Advanced Analysis Tab
+        with gr.Tab("Advanced Analysis"):
+            gr.Markdown("## 🔬 Explainable AI Analysis")
+            gr.Markdown("Use SHAP and LIME to understand which words and phrases most influence the sentiment prediction.")
             
-            with gr.Row():
-                batch_plot = gr.Plot(label="Batch Analysis Summary")
-                confidence_dist_plot = gr.Plot(label="Confidence Distribution")
-        
-        with gr.Tab("History & Analytics"):
             with gr.Row():
                 with gr.Column():
-                    with gr.Row():
-                        refresh_history_btn = gr.Button("Refresh History")
-                        clear_history_btn = gr.Button("Clear History", variant="stop")
-                        status_btn = gr.Button("Get Status")
-                    
-                    history_theme = gr.Dropdown(
-                        choices=list(config.THEMES.keys()),
-                        value="default",
-                        label="Dashboard Theme"
+                    advanced_text_input = gr.Textbox(
+                        label="Enter Text for Advanced Analysis",
+                        placeholder="Enter text to analyze with SHAP and LIME...",
+                        lines=6
                     )
                     
-                    with gr.Row():
-                        export_csv_btn = gr.Button("Export CSV")
-                        export_json_btn = gr.Button("Export JSON")
-                
-                with gr.Column():
-                    history_status = gr.Textbox(label="History Status", lines=8)
-            
-            history_dashboard = gr.Plot(label="History Analytics Dashboard")
-            
-            with gr.Row():
-                csv_download = gr.File(label="CSV Download", visible=True)
-                json_download = gr.File(label="JSON Download", visible=True)
-        
-        # Event Handlers
-        analyze_btn.click(
-            app.analyze_single,
-            inputs=[text_input, language_selector, theme_selector, 
-                   clean_text_cb, remove_punct_cb, remove_nums_cb],
-            outputs=[result_output, gauge_plot, probability_plot, keyword_plot]
-        )
-        
-        load_file_btn.click(
-            app.data_handler.process_file,
-            inputs=file_upload,
-            outputs=batch_input
-        )
-        
-        analyze_batch_btn.click(
-            app.analyze_batch,
-            inputs=[batch_input, batch_language, batch_theme,
-                   batch_clean_cb, batch_punct_cb, batch_nums_cb],
-            outputs=[batch_summary, batch_results_df, batch_plot, confidence_dist_plot]
-        )
-        
-        refresh_history_btn.click(
-            app.plot_history,
-            inputs=history_theme,
-            outputs=[history_dashboard, history_status]
-        )
-        
-        clear_history_btn.click(
-            lambda: f"Cleared {app.history.clear()} entries",
-            outputs=history_status
-        )
-        
-        status_btn.click(
-            app.get_history_status,
-            outputs=history_status
-        )
-        
-        export_csv_btn.click(
-            lambda: app.data_handler.export_data(app.history.get_all(), 'csv'),
-            outputs=[csv_download, history_status]
-        )
-        
-        export_json_btn.click(
-            lambda: app.data_handler.export_data(app.history.get_all(), 'json'),
-            outputs=[json_download, history_status]
-        )
-    
-    return demo
-
-# Application Entry Point
-if __name__ == "__main__":
-    logging.basicConfig(
-        level=logging.INFO,
-        format='%(asctime)s - %(name)s - %(levelname)s - %(message)s'
-    )
-    
-    try:
-        demo = create_interface()
-        demo.launch(
-            share=True,
-            server_name="0.0.0.0",
-            server_port=7860,
-            show_error=True
-        )
-    except Exception as e:
-        logger.error(f"Failed to launch application: {e}")
-        raise
-    
-    @handle_errors(default_return=("Please enter texts", None, None, None))
-    def analyze_batch(self, batch_text: str, language: str, theme: str, 
-                     clean_text: bool, remove_punct: bool, remove_nums: bool):
-        """Enhanced batch analysis"""
-        if not batch_text.strip():
-            return "Please enter texts (one per line)", None, None, None
-        
-        # Parse batch input
-        texts = TextProcessor.parse_batch_input(batch_text)
-        
-        if len(texts) > config.BATCH_SIZE_LIMIT:
-            return f"Too many texts. Maximum {config.BATCH_SIZE_LIMIT} allowed.", None, None, None
-        
-        if not texts:
-            return "No valid texts found", None, None, None
-
-        
-        # Map display names to language codes
-        language_map = {v: k for k, v in config.SUPPORTED_LANGUAGES.items()}
-        language_code = language_map.get(language, 'auto')
-        
-        preprocessing_options = {
-            'clean_text': clean_text,
-            'remove_punctuation': remove_punct,
-            'remove_numbers': remove_nums
-        }
-        
-        with memory_cleanup():
-            results = self.engine.analyze_batch(texts, language_code, preprocessing_options)
-            
-            # Add to history
-            batch_entries = []
-            for result in results:
-                if 'error' not in result:
-                    entry = {
-                        'text': result['text'],
-                        'full_text': result['full_text'],
-                        'sentiment': result['sentiment'],
-                        'confidence': result['confidence'],
-                        'pos_prob': result.get('pos_prob', 0),
-                        'neg_prob': result.get('neg_prob', 0),
-                        'neu_prob': result.get('neu_prob', 0),
-                        'language': result['language'],
-                        'keywords': result['keywords'],
-                        'word_count': result['word_count'],
-                        'analysis_type': 'batch',
-                        'batch_index': result['batch_index']
-                    }
-                    batch_entries.append(entry)
-            
-            self.history.add_batch(batch_entries)
-            
-            # Create visualizations
-            theme_ctx = ThemeContext(theme)
-            summary_fig = PlotlyVisualizer.create_batch_summary(results, theme_ctx)
-            confidence_fig = PlotlyVisualizer.create_confidence_distribution(results)
-            
-            # Create results DataFrame
-            df_data = []
-            for result in results:
-                if 'error' in result:
-                    df_data.append({
-                        'Index': result['batch_index'] + 1,
-                        'Text': result['text'],
-                        'Sentiment': 'Error',
-                        'Confidence': 0.0,
-                        'Language': 'Unknown',
-                        'Error': result['error']
-                    })
-                else:
-                    keywords_str = ', '.join([word for word, _ in result['keywords'][:3]])
-                    df_data.append({
-                        'Index': result['batch_index'] + 1,
-                        'Text': result['text'],
-                        'Sentiment': result['sentiment'],
-                        'Confidence': f"{result['confidence']:.3f}",
-                        'Language': result['language'].upper(),
-                        'Keywords': keywords_str
-                    })
-            
-            df = pd.DataFrame(df_data)
-            
-            # Create summary text
-            successful_results = [r for r in results if 'error' not in r]
-            error_count = len(results) - len(successful_results)
-            
-            if successful_results:
-                sentiment_counts = Counter([r['sentiment'] for r in successful_results])
-                avg_confidence = np.mean([r['confidence'] for r in successful_results])
-                languages = Counter([r['language'] for r in successful_results])
-                
-                summary_text = f"""
-**Batch Analysis Summary:**
-- **Total Texts:** {len(texts)}
-- **Successful:** {len(successful_results)}
-- **Errors:** {error_count}
-- **Average Confidence:** {avg_confidence:.3f}
-- **Sentiments:** {dict(sentiment_counts)}
-- **Languages Detected:** {dict(languages)}
-                """
-            else:
-                summary_text = f"All {len(texts)} texts failed to analyze."
-            
-            return summary_text, df, summary_fig, confidence_fig
-    
-    @handle_errors(default_return=(None, "No history available"))
-    def plot_history(self, theme: str = 'default'):
-        """Plot comprehensive history analysis"""
-        history = self.history.get_all()
-        if len(history) < 2:
-            return None, f"Need at least 2 analyses for trends. Current: {len(history)}"
-        
-        theme_ctx = ThemeContext(theme)
-        
-        with memory_cleanup():
-            fig = PlotlyVisualizer.create_history_dashboard(history, theme_ctx)
-            stats = self.history.get_stats()
-            
-            stats_text = f"""
-**History Statistics:**
-- **Total Analyses:** {stats.get('total_analyses', 0)}
-- **Positive:** {stats.get('positive_count', 0)}
-- **Negative:** {stats.get('negative_count', 0)}
-- **Neutral:** {stats.get('neutral_count', 0)}
-- **Average Confidence:** {stats.get('avg_confidence', 0):.3f}
-- **Languages:** {stats.get('languages_detected', 0)}
-- **Most Common Language:** {stats.get('most_common_language', 'N/A').upper()}
-            """
-            
-            return fig, stats_text
-    
-    @handle_errors(default_return=("No data available",))
-    def get_history_status(self):
-        """Get current history status"""
-        stats = self.history.get_stats()
-        if not stats:
-            return "No analyses performed yet"
-        
-        return f"""
-**Current Status:**
-- **Total Analyses:** {stats['total_analyses']}
-- **Recent Sentiment Distribution:** 
-  * Positive: {stats['positive_count']}
-  * Negative: {stats['negative_count']}  
-  * Neutral: {stats['neutral_count']}
-- **Average Confidence:** {stats['avg_confidence']:.3f}
-- **Languages Detected:** {stats['languages_detected']}
-        """
-
-# Gradio Interface
-def create_interface():
-    """Create comprehensive Gradio interface"""
-    app = SentimentApp()
-    
-    with gr.Blocks(theme=gr.themes.Soft(), title="Multilingual Sentiment Analyzer") as demo:
-        gr.Markdown("# 🌍 Advanced Multilingual Sentiment Analyzer")
-        gr.Markdown("AI-powered sentiment analysis with support for multiple languages, advanced visualizations, and explainable AI features")
-        
-        with gr.Tab("Single Analysis"):
-            with gr.Row():
-                with gr.Column():
-                    text_input = gr.Textbox(
-                        label="Enter Text for Analysis",
-                        placeholder="Enter your text in any supported language...",
-                        lines=5
+                    advanced_language = gr.Dropdown(
+                        choices=list(config.SUPPORTED_LANGUAGES.values()),
+                        value="Auto Detect",
+                        label="Language"
                     )
                     
                     with gr.Row():
-                        language_selector = gr.Dropdown(
-                            choices=list(config.SUPPORTED_LANGUAGES.values()),
-                            value="Auto Detect",
-                            label="Language"
-                        )
-                        theme_selector = gr.Dropdown(
-                            choices=list(config.THEMES.keys()),
-                            value="default",
-                            label="Theme"
-                        )
-                    
-                    with gr.Row():
-                        clean_text_cb = gr.Checkbox(label="Clean Text", value=False)
-                        remove_punct_cb = gr.Checkbox(label="Remove Punctuation", value=False)
-                        remove_nums_cb = gr.Checkbox(label="Remove Numbers", value=False)
+                        shap_btn = gr.Button("SHAP Analysis", variant="primary")
+                        lime_btn = gr.Button("LIME Analysis", variant="secondary")
                     
-                    analyze_btn = gr.Button("Analyze", variant="primary", size="lg")
-                    
-                    gr.Examples(
-                        examples=app.examples,
-                        inputs=text_input,
-                        cache_examples=False
-                    )
+                    gr.Markdown("""
+                    **Analysis Methods:**
+                    - **SHAP**: Shows token-level importance scores
+                    - **LIME**: Explains predictions by perturbing input features
+                    """)
                 
                 with gr.Column():
-                    result_output = gr.Textbox(label="Analysis Results", lines=8)
+                    advanced_results = gr.Textbox(label="Analysis Summary", lines=10)
             
             with gr.Row():
-                gauge_plot = gr.Plot(label="Sentiment Gauge")
-                probability_plot = gr.Plot(label="Probability Distribution")
-            
-            with gr.Row():
-
-                keyword_plot = gr.Plot(label="Key Contributing Words")
+                advanced_plot = gr.Plot(label="Feature Importance Visualization")
         
         with gr.Tab("Batch Analysis"):
             with gr.Row():
@@ -1481,13 +1280,10 @@ def create_interface():
             with gr.Row():
                 csv_download = gr.File(label="CSV Download", visible=True)
                 json_download = gr.File(label="JSON Download", visible=True)
-
-
-
-
-
         
         # Event Handlers
+        
+        # Single Analysis
         analyze_btn.click(
             app.analyze_single,
             inputs=[text_input, language_selector, theme_selector, 
@@ -1495,6 +1291,20 @@ def create_interface():
             outputs=[result_output, gauge_plot, probability_plot, keyword_plot]
         )
         
+        # Advanced Analysis (NEW)
+        shap_btn.click(
+            app.analyze_with_shap,
+            inputs=[advanced_text_input, advanced_language],
+            outputs=[advanced_results, advanced_plot]
+        )
+        
+        lime_btn.click(
+            app.analyze_with_lime,
+            inputs=[advanced_text_input, advanced_language],
+            outputs=[advanced_results, advanced_plot]
+        )
+        
+        # Batch Analysis
         load_file_btn.click(
             app.data_handler.process_file,
             inputs=file_upload,
@@ -1508,6 +1318,7 @@ def create_interface():
             outputs=[batch_summary, batch_results_df, batch_plot, confidence_dist_plot]
         )
         
+        # History & Analytics
         refresh_history_btn.click(
             app.plot_history,
             inputs=history_theme,