Create app.py

app.py

@@ -0,0 +1,46 @@
+import gradio as gr
+from sentence_transformers import SentenceTransformer, util
+from transformers import pipeline
+
+# --- MODELS ---
+embedder = SentenceTransformer("all-MiniLM-L6-v2")
+llm = pipeline("text-generation", model="gpt2", max_new_tokens=60)
+
+# --- RULE-BASED FUNCTION ---
+def rule_based(text):
+    if "anaemia" in text.lower() or "anemia" in text.lower():
+        return "✅ Anaemia detected (keyword match)"
+    else:
+        return "❌ Anaemia not detected (no keyword match)"
+
+# --- MACHINE LEARNING FUNCTION ---
+reference_text = "Patient diagnosed with anaemia and low haemoglobin."
+ref_emb = embedder.encode(reference_text, convert_to_tensor=True)
+
+def ml_based(text):
+    emb = embedder.encode(text, convert_to_tensor=True)
+    score = util.cos_sim(emb, ref_emb).item()
+    return f"Similarity to anaemia reference: {score:.2f}\n" + \
+           ("✅ Anaemia likely" if score > 0.45 else "❌ Anaemia unlikely")
+
+# --- AI / FOUNDATION MODEL FUNCTION ---
+def ai_based(text):
+    prompt = f"Determine if the following note suggests anaemia:\n\n{text}\n\nAnswer clearly with reasoning."
+    result = llm(prompt)[0]["generated_text"]
+    return result
+
+# --- GRADIO UI ---
+with gr.Blocks(title="Rule-Based vs Machine Learning vs AI") as demo:
+    gr.Markdown("# 🔍 Clinical Text Understanding Demo\nCompare rule-based, ML, and AI approaches for detecting **anaemia**.")
+    text_input = gr.Textbox(label="Enter clinical note", lines=4)
+    rb_output = gr.Textbox(label="Rule-Based Result")
+    ml_output = gr.Textbox(label="Machine Learning Result")
+    ai_output = gr.Textbox(label="AI (LLM) Result")
+
+    def process(text):
+        return rule_based(text), ml_based(text), ai_based(text)
+
+    btn = gr.Button("Run All")
+    btn.click(process, inputs=text_input, outputs=[rb_output, ml_output, ai_output])
+
+demo.launch()