Update app.py

app.py

@@ -3,35 +3,41 @@ import torch
 from transformers import AutoModelForSequenceClassification, AutoTokenizer
 import requests
 import re
+import numpy as np
 from bs4 import BeautifulSoup
+from datasets import load_dataset
 
-# Hugging Face AI Model
+# Load AI detection model
 MODEL_NAME = "roberta-base-openai-detector"
 tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
 model = AutoModelForSequenceClassification.from_pretrained(MODEL_NAME)
 
-# AI Content Detection Function
+# AI content detection function
 def detect_ai_content(text):
     inputs = tokenizer(text, return_tensors="pt", truncation=True, padding=True)
     outputs = model(**inputs)
     scores = torch.nn.functional.softmax(outputs.logits, dim=1)
-    return scores[0][1].item()
+    return scores[0][1].item()  # AI probability
 
-# Writing Style Analysis
+# Writing style analysis
 def stylometry_analysis(text):
     words = text.split()
     avg_word_length = sum(len(word) for word in words) / len(words) if words else 0
     complex_words_ratio = len([word for word in words if len(word) > 6]) / len(words) if words else 0
     passive_voice_count = len(re.findall(r'\b(is|was|were|has been|have been|had been)\b \w+ed', text))
-    return {"avg_word_length": avg_word_length, "complex_words_ratio": complex_words_ratio, "passive_voice_count": passive_voice_count}
+    return {
+        "avg_word_length": avg_word_length,
+        "complex_words_ratio": complex_words_ratio,
+        "passive_voice_count": passive_voice_count
+    }
 
-# Semantic Similarity Analysis
+# Semantic similarity analysis
 def semantic_analysis(text):
     keywords = ["AI", "generated", "neural network", "LLM", "GPT", "transformer"]
     ai_patterns = sum([text.lower().count(keyword) for keyword in keywords])
     return ai_patterns / len(text.split()) if text.split() else 0
 
-# Web Crawling for Plagiarism
+# Web crawling for plagiarism detection
 def check_plagiarism(text):
     search_url = f"https://www.google.com/search?q={'+'.join(text.split()[:10])}"
     headers = {"User-Agent": "Mozilla/5.0"}
@@ -42,24 +48,53 @@ def check_plagiarism(text):
         return any("wikipedia" in link or "researchgate" in link or "arxiv" in link for link in links)
     return False
 
+# Self-learning system (Internet AI Content Collection)
+def collect_ai_generated_text():
+    dataset = load_dataset("openai/webgpt_comparisons", split="train")
+    ai_text_samples = [item["answer_1"] for item in dataset.select(range(50))]
+    return ai_text_samples
+
+# Auto-learn from new AI models
+def update_model_with_new_patterns():
+    new_ai_texts = collect_ai_generated_text()
+    for text in new_ai_texts:
+        ai_score = detect_ai_content(text)
+        if ai_score < 0.5:
+            # Fine-tune model logic (this will need cloud-based model retraining)
+            pass
+
 # Streamlit UI
-st.title("🚀 Ultra-Advanced AI Content Detector")
-st.write("🔍 Detect whether the text is AI-generated or human-written.")
+st.title("Self-Learning AI Content Detector")
+st.write("Detect AI-generated text and continuously learn from new AI models.")
+
+text_input = st.text_area("Enter text to analyze:")
 
-text_input = st.text_area("📝 Enter Text Below:", "")
+if st.button("Analyze"):
+    if text_input.strip():
+        ai_probability = detect_ai_content(text_input)
+        writing_features = stylometry_analysis(text_input)
+        semantic_score = semantic_analysis(text_input)
+        is_plagiarized = check_plagiarism(text_input)
 
-if st.button("Analyze Text"):
-    ai_probability = detect_ai_content(text_input)
-    writing_features = stylometry_analysis(text_input)
-    semantic_score = semantic_analysis(text_input)
-    is_plagiarized = check_plagiarism(text_input)
+        # Multi-layer AI detection logic
+        is_ai_generated = (
+            ai_probability > 0.5 or 
+            writing_features["complex_words_ratio"] > 0.4 or 
+            semantic_score > 0.2 or 
+            is_plagiarized
+        )
 
-    is_ai_generated = (ai_probability > 0.5 or writing_features["complex_words_ratio"] > 0.4 or semantic_score > 0.2 or is_plagiarized)
+        result = "AI-Generated" if is_ai_generated else "Human-Written"
 
-    st.subheader("📊 Analysis Results:")
-    st.write(f"🔹 AI Probability: {ai_probability:.2f}")
-    st.write(f"🔹 Writing Features: {writing_features}")
-    st.write(f"🔹 Semantic Score: {semantic_score:.2f}")
-    st.write(f"🔹 Plagiarism Detected: {'Yes' if is_plagiarized else 'No'}")
-    st.subheader("🧐 Final Verdict:")
-    st.write("✅ **Human-Written**" if not is_ai_generated else "❌ **AI-Generated**")
+        st.subheader("Detection Results")
+        st.write(f"**AI Probability:** {ai_probability:.2f}")
+        st.write(f"**Complex Words Ratio:** {writing_features['complex_words_ratio']:.2f}")
+        st.write(f"**Passive Voice Count:** {writing_features['passive_voice_count']}")
+        st.write(f"**Semantic Score:** {semantic_score:.2f}")
+        st.write(f"**Plagiarism Detected:** {'Yes' if is_plagiarized else 'No'}")
+        st.subheader(f"Final Verdict: {result}")
+        
+        # Auto-learn from new AI patterns
+        update_model_with_new_patterns()
+    else:
+        st.error("Please enter some text for analysis.")