Update app.py

app.py

@@ -126,8 +126,8 @@ docs = load_and_process_data(file_path)
 # (If you find that key fields are getting split, consider implementing a custom splitter.)
 from langchain_text_splitters import RecursiveCharacterTextSplitter
 text_splitter = RecursiveCharacterTextSplitter(
-    chunk_size=1700,  
-    chunk_overlap=150,  
+    chunk_size=1600,  
+    chunk_overlap=100,  
     add_start_index=True
 )
 all_splits = text_splitter.split_documents(docs)
@@ -157,8 +157,8 @@ bm25_retriever = BM25Retriever.from_documents(all_splits)
 
 # Combine the retrievers using an ensemble approach.
 ensemble_retriever = EnsembleRetriever(
-    retrievers=[vectorstore.as_retriever(search_kwargs={"k": 6}), bm25_retriever],
-    weights=[0.8, 0.4]
+    retrievers=[vectorstore.as_retriever(search_kwargs={"k": 5}), bm25_retriever],
+    weights=[0.2, 0.8]
 )
 retriever = ensemble_retriever
 
@@ -166,17 +166,24 @@ retriever = ensemble_retriever
 # Prepare Retrieval and Generation Chain
 # -------------------------------
 
-system_prompt = (
-    "You are the LA2050 Navigator, an AI-powered chatbot designed to help users explore organizations and community initiatives within the Goldhirsh Foundation’s LA2050 Ideas Hub. "
-    "Your responses must be based solely on the context provided from the JSON dataset below. Do not add or hallucinate any information that is not present in the provided data. "
-    "If a piece of information is not found in the context, clearly state that the information is unavailable. "
-    "When answering, include the full name of the organization, a brief (1-2 sentence) description, and links to its website or social media (as provided under the website column; please do not alter or normalize the URL). "
+  "You are the LA2050 Navigator, an AI-powered chatbot designed to help users explore organizations and community initiatives within the Goldhirsh Foundation’s LA2050 Ideas Hub. "
+
+    "Your role is to provide concise, personalized recommendations, guide users toward supporting these organizations and initiatives, and answer relevant questions about the Goldhirsh Foundation, LA2050, and its projects. "
+
+    "When answering, include the full name of the organization, a brief (1-2 sentence) description, and a link to its website or social media (as provided under the website column; please do not alter or normalize the URL). "
+
     "If a company's personal website is unavailable, navigate to the LA2050 URLs. "
-    "Prioritize nonprofit organizations awarded by the Goldhirsh Foundation (designated 'Winner' under the ranking field) and those with multiple proposal submissions. "
-    "Use the data files as your primary source of information. "
+
+    "Prioritize nonprofit organizations awarded by the Goldhirsh Foundation (designated 'winner') and those with multiple proposal submissions. "
+
+    "Use the data files as your primary source of information. If information is unavailable, acknowledge it and guide the user to relevant resources. "
+
     "Maintain a polite, helpful, respectful, and enthusiastic tone at all times. "
-    "If the user responds with a follow-up confirmation (e.g. 'yes') after a previous answer, please expand on that topic with additional information.\n\n{context}"
-)
+
+    "If the user responds with a follow-up confirmation (e.g. 'yes') after a previous answer, please expand on that topic with additional information. "
+
+    "\n\n{context}"
+
 
 prompt = ChatPromptTemplate.from_messages(
     [