tuesday openai

app.py

@@ -5,6 +5,11 @@ from transformers import pipeline
 import re
 from langdetect import detect
 from deep_translator import GoogleTranslator
+import openai
+import os
+
+# Set up your OpenAI API key
+openai.api_key = os.getenv("OPENAI_API_KEY")
 
 # Translator instance
 translator = GoogleTranslator(source="auto", target="es")
@@ -23,7 +28,7 @@ CANDIDATE_LABELS = ["SMiShing", "Other Scam", "Legitimate"]
 
 def get_keywords_by_language(text: str):
     """
-    Detect language using `langdetect` and translate keywords if needed.
+    Detect language using langdetect and translate keywords if needed.
     """
     snippet = text[:200]
     try:
@@ -88,9 +93,129 @@ def boost_probabilities(probabilities: dict, text: str):
         "detected_lang": detected_lang
     }
 
+def query_llm_for_classification(raw_message: str) -> dict:
+    """
+    First LLM call: asks for a classification (SMiShing, Other Scam, or Legitimate)
+    acting as a cybersecurity expert. Returns label and short reason.
+    """
+    if not raw_message.strip():
+        return {"label": "Unknown", "reason": "No message provided to the LLM."}
+
+    system_prompt = (
+        "You are a cybersecurity expert. You will classify the user's message "
+        "as one of: SMiShing, Other Scam, or Legitimate. Provide a short reason. "
+        "Return only JSON with keys: label, reason."
+    )
+    user_prompt = f"Message: {raw_message}\nClassify it as SMiShing, Other Scam, or Legitimate."
+
+    try:
+        response = openai.ChatCompletion.create(
+            model="gpt-3.5-turbo",
+            messages=[
+                {"role": "system", "content": system_prompt},
+                {"role": "user", "content": user_prompt}
+            ],
+            temperature=0.2
+        )
+        raw_reply = response["choices"][0]["message"]["content"].strip()
+
+        import json
+        # Expect something like {"label": "...", "reason": "..."}
+        llm_result = json.loads(raw_reply)
+        if "label" not in llm_result or "reason" not in llm_result:
+            return {"label": "Unknown", "reason": f"Unexpected format: {raw_reply}"}
+
+        return llm_result
+
+    except Exception as e:
+        return {"label": "Unknown", "reason": f"LLM error: {e}"}
+
+def incorporate_llm_label(boosted: dict, llm_label: str) -> dict:
+    """
+    Adjust the final probabilities based on the LLM's classification.
+    If LLM says SMiShing, add +0.2 to SMiShing, etc. Then clamp & re-normalize.
+    """
+    if llm_label == "SMiShing":
+        boosted["SMiShing"] += 0.2
+    elif llm_label == "Other Scam":
+        boosted["Other Scam"] += 0.2
+    elif llm_label == "Legitimate":
+        boosted["Legitimate"] += 0.2
+    # else "Unknown" => do nothing
+
+    # clamp
+    for k in boosted:
+        if boosted[k] < 0:
+            boosted[k] = 0.0
+
+    total = sum(boosted.values())
+    if total > 0:
+        for k in boosted:
+            boosted[k] /= total
+    else:
+        # fallback
+        boosted["Legitimate"] = 1.0
+        boosted["SMiShing"] = 0.0
+        boosted["Other Scam"] = 0.0
+
+    return boosted
+
+def query_llm_for_explanation(
+    text: str,
+    final_label: str,
+    final_conf: float,
+    local_label: str,
+    local_conf: float,
+    llm_label: str,
+    llm_reason: str,
+    found_smishing: list,
+    found_other_scam: list,
+    found_urls: list
+) -> str:
+    """
+    Second LLM call: provides a holistic explanation of the final classification.
+    We include the local classification info, the LLM’s own classification, and
+    relevant details (keywords, URLs).
+    """
+    system_prompt = (
+        "You are a cybersecurity expert providing a final explanation to the user. "
+        "Combine the local classification, the LLM classification, and the final label "
+        "into one concise explanation. Do not reveal internal code or raw JSON; just give "
+        "a short, user-friendly explanation. End with a final statement of the final label."
+    )
+    user_context = f"""
+User Message:
+{text}
+
+Local Classification => Label: {local_label}, Confidence: {local_conf}
+LLM Classification => Label: {llm_label}, Reason: {llm_reason}
+Final Overall Label => {final_label} (confidence {final_conf})
+
+Suspicious SMiShing Keywords => {found_smishing}
+Suspicious Other Scam Keywords => {found_other_scam}
+URLs => {found_urls}
+"""
+    # The LLM can combine these facts into a short paragraph.
+    try:
+        response = openai.ChatCompletion.create(
+            model="gpt-3.5-turbo",
+            messages=[
+                {"role": "system", "content": system_prompt},
+                {"role": "user", "content": user_context}
+            ],
+            temperature=0.2
+        )
+        final_explanation = response["choices"][0]["message"]["content"].strip()
+        return final_explanation
+    except Exception as e:
+        return f"Could not generate final explanation due to error: {e}"
+
 def smishing_detector(input_type, text, image):
     """
-    Main detection function combining text (if 'Text') and OCR (if 'Screenshot').
+    Main detection function combining text (if 'Text') & OCR (if 'Screenshot'),
+    plus two LLM calls: 
+      1) classification to adjust final probabilities,
+      2) a final explanation summarizing the outcome.
     """
     if input_type == "Text":
         combined_text = text.strip() if text else ""
@@ -105,31 +230,44 @@ def smishing_detector(input_type, text, image):
             "label": "No text provided",
             "confidence": 0.0,
             "keywords_found": [],
-            "urls_found": []
+            "urls_found": [],
+            "llm_label": "Unknown",
+            "llm_reason": "No text to analyze",
+            "final_explanation": "No text provided"
         }
 
-    result = classifier(
+    # 1. Local zero-shot classification
+    local_result = classifier(
         sequences=combined_text,
         candidate_labels=CANDIDATE_LABELS,
         hypothesis_template="This message is {}."
     )
-    original_probs = {k: float(v) for k, v in zip(result["labels"], result["scores"])}
+    original_probs = {k: float(v) for k, v in zip(local_result["labels"], local_result["scores"])}
 
+    # 2. Basic boosting from keywords & URLs
     boosted = boost_probabilities(original_probs, combined_text)
+    detected_lang = boosted.pop("detected_lang", "en")
+
+    # Convert to float only
+    for k in boosted:
+        boosted[k] = float(boosted[k])
+
+    local_label = max(boosted, key=boosted.get)
+    local_conf = round(boosted[local_label], 3)
 
-    # Patched snippet begins
-    # 1. Extract language first, preserving it
-    detected_lang = boosted.get("detected_lang", "en")
-    # 2. Remove it so only numeric keys remain
-    boosted.pop("detected_lang", None)
-    # 3. Convert numeric values to float
-    for k, v in boosted.items():
-        boosted[k] = float(v)
-    # Patched snippet ends
+    # 3. LLM Classification
+    llm_classification = query_llm_for_classification(combined_text)
+    llm_label = llm_classification.get("label", "Unknown")
+    llm_reason = llm_classification.get("reason", "No reason provided")
 
+    # 4. Incorporate LLM’s label into final probabilities
+    boosted = incorporate_llm_label(boosted, llm_label)
+
+    # Now we have updated probabilities
     final_label = max(boosted, key=boosted.get)
     final_confidence = round(boosted[final_label], 3)
 
+    # 5. Gather found keywords & URLs
     lower_text = combined_text.lower()
     smishing_keys, scam_keys, _ = get_keywords_by_language(combined_text)
 
@@ -137,16 +275,34 @@ def smishing_detector(input_type, text, image):
     found_smishing = [kw for kw in smishing_keys if kw in lower_text]
     found_other_scam = [kw for kw in scam_keys if kw in lower_text]
 
+    # 6. Final LLM explanation
+    final_explanation = query_llm_for_explanation(
+        text=combined_text,
+        final_label=final_label,
+        final_conf=final_confidence,
+        local_label=local_label,
+        local_conf=local_conf,
+        llm_label=llm_label,
+        llm_reason=llm_reason,
+        found_smishing=found_smishing,
+        found_other_scam=found_other_scam,
+        found_urls=found_urls
+    )
+
     return {
         "detected_language": detected_lang,
         "text_used_for_classification": combined_text,
         "original_probabilities": {k: round(v, 3) for k, v in original_probs.items()},
-        "boosted_probabilities": {k: round(v, 3) for k, v in boosted.items()},
+        "boosted_probabilities_before_llm": {local_label: local_conf},
+        "llm_label": llm_label,
+        "llm_reason": llm_reason,
+        "boosted_probabilities_after_llm": {k: round(v, 3) for k, v in boosted.items()},
         "label": final_label,
         "confidence": final_confidence,
         "smishing_keywords_found": found_smishing,
         "other_scam_keywords_found": found_other_scam,
         "urls_found": found_urls,
+        "final_explanation": final_explanation,
     }
 
 #
@@ -165,7 +321,7 @@ def toggle_inputs(choice):
         return gr.update(visible=False), gr.update(visible=True)
 
 with gr.Blocks() as demo:
-    gr.Markdown("## SMiShing & Scam Detector (Choose Text or Screenshot)")
+    gr.Markdown("## SMiShing & Scam Detector with LLM-Enhanced Logic")
     
     with gr.Row():
         input_type = gr.Radio(
@@ -207,4 +363,7 @@ with gr.Blocks() as demo:
     )
 
 if __name__ == "__main__":
+    # Warn if openai.api_key not set
+    if not openai.api_key:
+        print("WARNING: OPENAI_API_KEY not set. LLM calls will fail or be skipped.")
     demo.launch()

app.py.bestoftues

@@ -0,0 +1,210 @@
+import gradio as gr
+import pytesseract
+from PIL import Image
+from transformers import pipeline
+import re
+from langdetect import detect
+from deep_translator import GoogleTranslator
+
+# Translator instance
+translator = GoogleTranslator(source="auto", target="es")
+
+# 1. Load separate keywords for SMiShing and Other Scam (assumed in English)
+with open("smishing_keywords.txt", "r", encoding="utf-8") as f:
+    SMISHING_KEYWORDS = [line.strip().lower() for line in f if line.strip()]
+
+with open("other_scam_keywords.txt", "r", encoding="utf-8") as f:
+    OTHER_SCAM_KEYWORDS = [line.strip().lower() for line in f if line.strip()]
+
+# 2. Zero-Shot Classification Pipeline
+model_name = "joeddav/xlm-roberta-large-xnli"
+classifier = pipeline("zero-shot-classification", model=model_name)
+CANDIDATE_LABELS = ["SMiShing", "Other Scam", "Legitimate"]
+
+def get_keywords_by_language(text: str):
+    """
+    Detect language using `langdetect` and translate keywords if needed.
+    """
+    snippet = text[:200]
+    try:
+        detected_lang = detect(snippet)
+    except Exception:
+        detected_lang = "en"
+
+    if detected_lang == "es":
+        smishing_in_spanish = [
+            translator.translate(kw).lower() for kw in SMISHING_KEYWORDS
+        ]
+        other_scam_in_spanish = [
+            translator.translate(kw).lower() for kw in OTHER_SCAM_KEYWORDS
+        ]
+        return smishing_in_spanish, other_scam_in_spanish, "es"
+    else:
+        return SMISHING_KEYWORDS, OTHER_SCAM_KEYWORDS, "en"
+
+def boost_probabilities(probabilities: dict, text: str):
+    """
+    Boost probabilities based on keyword matches and presence of URLs.
+    """
+    lower_text = text.lower()
+    smishing_keywords, other_scam_keywords, detected_lang = get_keywords_by_language(text)
+
+    smishing_count = sum(1 for kw in smishing_keywords if kw in lower_text)
+    other_scam_count = sum(1 for kw in other_scam_keywords if kw in lower_text)
+
+    smishing_boost = 0.30 * smishing_count
+    other_scam_boost = 0.30 * other_scam_count
+
+    found_urls = re.findall(r"(https?://[^\s]+)", lower_text)
+    if found_urls:
+        smishing_boost += 0.35
+
+    p_smishing = probabilities.get("SMiShing", 0.0)
+    p_other_scam = probabilities.get("Other Scam", 0.0)
+    p_legit = probabilities.get("Legitimate", 1.0)
+
+    p_smishing += smishing_boost
+    p_other_scam += other_scam_boost
+    p_legit -= (smishing_boost + other_scam_boost)
+
+    # Clamp
+    p_smishing = max(p_smishing, 0.0)
+    p_other_scam = max(p_other_scam, 0.0)
+    p_legit = max(p_legit, 0.0)
+
+    # Re-normalize
+    total = p_smishing + p_other_scam + p_legit
+    if total > 0:
+        p_smishing /= total
+        p_other_scam /= total
+        p_legit /= total
+    else:
+        p_smishing, p_other_scam, p_legit = 0.0, 0.0, 1.0
+
+    return {
+        "SMiShing": p_smishing,
+        "Other Scam": p_other_scam,
+        "Legitimate": p_legit,
+        "detected_lang": detected_lang
+    }
+
+def smishing_detector(input_type, text, image):
+    """
+    Main detection function combining text (if 'Text') and OCR (if 'Screenshot').
+    """
+    if input_type == "Text":
+        combined_text = text.strip() if text else ""
+    else:
+        combined_text = ""
+        if image is not None:
+            combined_text = pytesseract.image_to_string(image, lang="spa+eng").strip()
+
+    if not combined_text:
+        return {
+            "text_used_for_classification": "(none)",
+            "label": "No text provided",
+            "confidence": 0.0,
+            "keywords_found": [],
+            "urls_found": []
+        }
+
+    result = classifier(
+        sequences=combined_text,
+        candidate_labels=CANDIDATE_LABELS,
+        hypothesis_template="This message is {}."
+    )
+    original_probs = {k: float(v) for k, v in zip(result["labels"], result["scores"])}
+
+    boosted = boost_probabilities(original_probs, combined_text)
+
+    # Patched snippet begins
+    # 1. Extract language first, preserving it
+    detected_lang = boosted.get("detected_lang", "en")
+    # 2. Remove it so only numeric keys remain
+    boosted.pop("detected_lang", None)
+    # 3. Convert numeric values to float
+    for k, v in boosted.items():
+        boosted[k] = float(v)
+    # Patched snippet ends
+
+    final_label = max(boosted, key=boosted.get)
+    final_confidence = round(boosted[final_label], 3)
+
+    lower_text = combined_text.lower()
+    smishing_keys, scam_keys, _ = get_keywords_by_language(combined_text)
+
+    found_urls = re.findall(r"(https?://[^\s]+)", lower_text)
+    found_smishing = [kw for kw in smishing_keys if kw in lower_text]
+    found_other_scam = [kw for kw in scam_keys if kw in lower_text]
+
+    return {
+        "detected_language": detected_lang,
+        "text_used_for_classification": combined_text,
+        "original_probabilities": {k: round(v, 3) for k, v in original_probs.items()},
+        "boosted_probabilities": {k: round(v, 3) for k, v in boosted.items()},
+        "label": final_label,
+        "confidence": final_confidence,
+        "smishing_keywords_found": found_smishing,
+        "other_scam_keywords_found": found_other_scam,
+        "urls_found": found_urls,
+    }
+
+#
+# Gradio interface with dynamic visibility
+#
+def toggle_inputs(choice):
+    """
+    Return updates for (text_input, image_input) based on the radio selection.
+    """
+    if choice == "Text":
+        # Show text input, hide image
+        return gr.update(visible=True), gr.update(visible=False)
+    else:
+        # choice == "Screenshot"
+        # Hide text input, show image
+        return gr.update(visible=False), gr.update(visible=True)
+
+with gr.Blocks() as demo:
+    gr.Markdown("## SMiShing & Scam Detector (Choose Text or Screenshot)")
+    
+    with gr.Row():
+        input_type = gr.Radio(
+            choices=["Text", "Screenshot"], 
+            value="Text", 
+            label="Choose Input Type"
+        )
+
+    text_input = gr.Textbox(
+        lines=3,
+        label="Paste Suspicious SMS Text",
+        placeholder="Type or paste the message here...",
+        visible=True  # default
+    )
+
+    image_input = gr.Image(
+        type="pil",
+        label="Upload Screenshot",
+        visible=False  # hidden by default
+    )
+
+    # Whenever input_type changes, toggle which input is visible
+    input_type.change(
+        fn=toggle_inputs,
+        inputs=input_type,
+        outputs=[text_input, image_input],
+        queue=False
+    )
+
+    # Button to run classification
+    analyze_btn = gr.Button("Classify")
+    output_json = gr.JSON(label="Result")
+
+    # On button click, call the smishing_detector
+    analyze_btn.click(
+        fn=smishing_detector,
+        inputs=[input_type, text_input, image_input],
+        outputs=output_json
+    )
+
+if __name__ == "__main__":
+    demo.launch()

requirements.txt

@@ -9,3 +9,4 @@ httpx==0.13.3
 sentencepiece==0.1.99
 numpy==1.25.0
 shap==0.41.0
+openai