Upload 2 files

app.py

@@ -0,0 +1,48 @@
+import gradio as gr
+from bertopic import BERTopic
+from pymongo import MongoClient
+
+# 🧠 Connect to MongoDB
+mongo_uri = "mongodb+srv://flaskuser:0000@cluster0.axlhy4c.mongodb.net/Researchersedst?retryWrites=true&w=majority"
+client = MongoClient(mongo_uri)
+
+
+db = client["Researchersedst"]
+collection = db["experiencesmk"]
+
+# 🧪 Topic Modeling Function
+def extract_topics_from_mongo(dummy_input=None):
+    # 🗃️ Fetch abstracts
+    articles = list(collection.find({"summary": {"$exists": True, "$ne": None}}, {"summary": 1, "_id": 0}))
+    texts = [article["summary"] for article in articles]
+
+    if not texts:
+        return "No abstracts found in MongoDB.", None
+
+    # 🧠 Run BERTopic
+    topic_model = BERTopic()
+    topics, _ = topic_model.fit_transform(texts)
+
+    # 📊 Topic Summary
+    topic_info = topic_model.get_topic_info()
+    top_topics = topic_info[['Topic', 'Name', 'Count']].head(10).to_string(index=False)
+
+    # 📎 Assign topics to texts
+    results = [{"topic": topic, "abstract": text} for topic, text in zip(topics, texts)]
+
+    return top_topics, results
+
+# 🎛️ Gradio UI
+demo = gr.Interface(
+    fn=extract_topics_from_mongo,
+    inputs=[],
+    outputs=[
+        gr.Textbox(label="Top Topics"),
+        gr.Dataframe(label="Abstracts with Assigned Topics")
+    ],
+    title="🧠 Topic Modeling from MongoDB",
+    description="This tool pulls abstracts from MongoDB and extracts topics using BERTopic."
+)
+
+if __name__ == "__main__":
+    demo.launch()

requirements.txt

@@ -0,0 +1,4 @@
+gradio
+bertopic
+pandas
+scikit-learn